Kathleen M' Turczyn, MPH

1
Measuring Disparity in Healthy People 2010
Physical Activity and Fitness Objectives

• Kathleen M. Turczyn, MPH
• Kenneth Keppel, PhD
• National Center for Health Statistics

APHA 131st Annual Meeting Monday, November 17,
2003
2
Healthy People 2010 goals

• 1. Increase quality and years of healthy life

• 2. Eliminate health disparities

among segments of the population, including
differences that occur by gender, race or
ethnicity, education or income, disability,
geographic location, or sexual orientation.
3
Definition of disparity
Disparity is the quantity that separates a group
from a reference point expressed in terms of a
rate, percent, mean, or other quantifiable
measure.
4
Physical Activity and FitnessGoal Improve
health, fitness, and quality of life through
daily physical activity

• Physical Activity in Adults
• 22-1 No leisure-time physical activity
• 22-2 Moderate physical activity
• 22-3 Vigorous physical activity
• Muscular Strength/Endurance and Flexibility
• 22-4 Muscular strength and endurance
• 22-5 Flexibility

• Physical Activity in Children and Adolescents
• 22-6 Moderate physical activity in adolescents
• 22-7 Vigorous physical activity in adolescents
• 22-8 Physical education requirement in schools
• 22-9 Daily physical education in schools
• 22-10 Physical activity in physical education
  class
• 22-11 Television viewing
• Access
• 22-12 School physical activity facilities
• 22-13 Worksite physical activity and fitness
• 22-14 Community walking
• 22-15 Community bicycling

5
HP2010 data template for population-based PAF
objectives

• Race and ethnicity
• Gender
• Education level (ages 25 years and over)
• Geographic location (adults)
• Disability status (adults)
• Select populations
• Age groups (adults)
• Persons with/without arthritis symptoms (adults)
• Grade level (adolescents)

6
Summary of attributes of Healthy People 2010
measure of disparity

• Disparity measured primarily in relative terms
• Disparity measure made on negative outcomes -
  objectives transposed, as necessary, to
  rates/percents of adverse events
• More favorable (best group) rate used as
  reference point
• Percent difference used between one or more
  individual groups and a common reference point
• Index of disparity used as a summary measure
  for a domain of several groups

7
Absolute vs. relative measures Adults
participating in vigorous physical activity, 2001
Percent
Female
Male
No vigorous activity
80
80-73 7
73
80-73/73 10
Do vigorous activity
20-27 7
20-27/27 26
27
20
No vigorous activity
Do vigorous activity
Source National Health Interview Survey, NCHS,
CDC.
8
Absolute vs. relative measures Adults
participating in vigorous physical activity, 2001
Percent
Female
Male
No vigorous activity
80
80-73 7
73
80-73/73 10
Relative
Do vigorous activity
20-27 7
Relative
20-27/27 26
27
20
No vigorous activity
Do vigorous activity
Source National Health Interview Survey, NCHS,
CDC.
9
Choice of reference point

• Target fixed no variability BUT cant
  distinguish between progress toward target and
  progress toward eliminating disparity
• Total population rate (weighted average of
  subgroup rates) reflects average experience of
  all groups in the population BUT influenced by
  change in distribution of the population among
  groups over time
• Mean of group rates (unweighted average of
  subgroup rates) not affected by changes in
  population distribution over time BUT more
  affected by outlying group rates than other
  measures
• Largest group rate reliable BUT singles out
  one group as the standard of comparison for all
  objectives
• Best group rate implicit is idea that all
  differences are inequitable (fits with goal of
  elimination of disparity) BUT differs from one
  domain to the next may change over time best
  group might be the least reliable

10
Different disparity measures for number of groups
in a domain

• Between 2 groups (one is best rate)
• Use the percent difference
• Among several groups (one is best rate)
• Use a summary measure the average of percent
  differences
• Index of disparity

11
Measuring disparity betweentwo groups

• Percent Difference Ri-Rr / Rr 100

Ri Rate or percent of specific group Rr Rate or
percent of reference population with best
rate or percent
12
Gender disparity for PAF objectives for adults 18
years and over
Change in disparity from baseline to most recent
(percentage points)
Prevalence
 
Most recent 2001
Baseline 1997/98
 
dif. from "best"
Rate or percent
dif. from "best"
Rate or percent
Objectives and Gender
 
 
 
 
 
 
 
 
Objective 22-01 1997 baseline
 
 
 
 
 
 
 
 


40.1

43.0
Female

 
 

0.0
34.7

0.0
36.2
Male
 
 
 
 
 
 
 
 
Objective 22-02 (T) 1997 baseline
 
 
 
 
 
 
 
 


71.2

81.3
Female

 
 

0.0
64.6

0.0
64.7
Male
 
 
 
 
 
 
 
 
Objective 22-03 (T) 1997 baseline
 
 
 
 
 
 
 
 


80.4

80.1
Female

 
 

0.0
73.0

0.0
73.4
Male
 
 
 
 
 
 
 
 
Objective 22-04 (T) 1997 baseline
 
 
 
 
 
 
 
 


82.8

85.6
Female

 
 

0.0
76.6

0.0
78.8
Male
 
 
 
 
 
 
 
 
Objective 22-05 (T) 1998 baseline
 
 
 
 
 
 
 
 


0.0
68.5

0.0
69.7
Female

 
 

70.4

70.3
Male
13
Measures of disparity between males and females
for selected PAF objectives, 2001
Adults
Adolescents
0
5
10
15
20
55
60
Percent difference from the best group
Ri-Rr / Rr100
Statistically significant at Plt0.05. Data for
adults are for ages 18 years and over data for
adolescents are for grades 9-12. Sources For
adults - CDC, NCHS, National Health Interview
Survey for adolescents CDC, NCCDPHP, Youth
Risk Behavior Surveillance System.
14
Genders with best rates for PAF objectives 2001

• Adolescents
• No moderate physical activity
• Male
• No vigorous physical activity
• Male
• No daily school PE
• Male
• Less than 50 active PE class time
• Male
• TV over 2 hours school days
• Female

• Adults
• No leisure-time physical activity
• Male
• No moderate physical activity
• Male
• No vigorous physical activity
• Male
• No strength/endurance exercises
• Male
• No stretching exercises for flexibility
• Female

15
Changes in disparity for males and females,
selected PAF objectives Baseline to most recent
update (2001)
More disparity over time
Less disparity over time
Adults
Adolescents
-20
-10
0
5
10
15
20
-15
-5
Percentage point change in disparity
Statistically significant at Plt0.05. Data for
adults are for ages 18 years and over data for
adolescents are for grades 9-12. Sources
Adults - CDC, NCHS, National Health Interview
Survey (1997/98-2001)
Adolescents CDC, NCCDPHP, Youth Risk Behavior
Surveillance System (1999-2001).
16
Measuring disparity among several groups

• Index of Disparity (Sn-1i1 PDi) / (n-1)

PDi Percent difference from best rate for
each group rate of interest n Total number of
groups
17
Race/ethnicity disparity for adults with no
leisure-time physical activity
Summary change in disparity from baseline
to most recent (percentage points)
Prevalence
 
Most recent 2001
Baseline for Race/Ethnic Groups
1999
 
dif. from "best"
Percent
dif. from "best"
Percent
Objectives and Race/Ethnicity
 
 
 
 
 
 
 
 
Objective 22-01
 
 
 
 
 
 
 
 
Adults with no leisure-time physical activity

18.0

59.8

41.9
American Indian or Alaska Native only
49.9
46.0

11.0

29.8

18.8
Asian only
40.5
38.5

-1.2

57.2

58.4
Black only
49.1
51.4

-2.2

13.4

15.6
White only
35.4
37.5

0.0

0.0

0.0
2 or more races
31.2
32.4

6.3

78.1

71.8
Hispanic or Latino
55.6
55.7

-2.5

13.2

15.7
Not Hispanic or Latino
35.3
37.5

-1.2

57.2

58.5
Black only
49.1
51.4

-3.9

5.2

9.2
White only

32.8
35.4
 
 
 
46.0
 
 
40.0
 
Index of disparity (percent)
6.0
Data are for adults ages 18 and over. Source
CDC, NCHS, National Health Interview Survey.
18
Measures of disparity for race/ethnicity for
selected PAF objectives, 2001
Adults (6 groups)
Adolescents (3 groups)
0
10
20
50
90
30
40
Index of disparity from the best group
Sn-1i1 PDi / n-1
Data for adults are for ages 18 years and over
data for adolescents are for grades 9-12.
Sources For adults - CDC, NCHS, National Health
Interview Survey for adolescents CDC, NCCDPHP,
Youth Risk Behavior Surveillance System.
19
Percent difference from best group rate for
race/ethnicity groups for no leisure-time
physical activity, adults 2001
Index of disparity 46.0
Best group Two or more races
Hispanic or Latino
Asian only
Black, not Hispanic only
American Indian/ Alaska Native only
White, not Hispanic only
0
40
60
20
80
Percent difference from the best group
Ri-Rr / Rr100
Data for adults are for ages 18 years and over.
Source CDC, NCHS, National Health Interview
Survey.
20
Percent difference from best group rate for
race/ethnicity groups for watching TV over 2
hours on school days, adolescents 2001
Index of disparity 88.2
Best group White, not Hispanic only
Black, not Hispanic only
Hispanic or Latino
0
40
60
20
80
120
100
140
Percent difference from the best group
Ri-Rr / Rr100
Data for adolescents are for grades 9-12.
Source CDC, NCCDPHP, Youth Risk Behavior
Surveillance System.
21
Race/ethnicity groups with the best rates for PAF
objectives 2001

• Adolescents
• No moderate physical activity
• White, not Hispanic only
• No vigorous physical activity
• White, not Hispanic only
• No daily school PE
• Hispanic or Latino
• Less than 50 active PE class time
• Hispanic or Latino
• TV over 2 hours school days
• White, not Hispanic only

• Adults
• No leisure-time physical activity
• Two or more races
• No moderate physical activity
• White, not Hispanic only
• No vigorous physical activity
• American Indian/Alaska Native
• No strength/endurance exercises
• Two or more races
• No stretching exercises for flexibility
• Asian only

22
Changes in disparity among race/ethnic groups,
selected PAF objectives 1999-2001
More disparity over time
Less disparity over time
Adults (6 groups)
Adolescents (3 groups)
Percentage point change in disparity
Data for adults are for ages 18 years and over
data for adolescents are for grades 9-12.
Sources Adults - CDC, NCHS, National Health
Interview Survey Adolescents CDC, NCCDPHP,
Youth Risk Behavior Surveillance System.
23
Measures of educational attainment disparity for
selected PAF objectives, 2001
Adults (5 groups)
Adolescents (4 grades)
0
10
20
30
40
130
50
60
Index of disparity from the best group
Sn-1i1 PDi / n-1
Index of disparity for flexibility is based on 3
education groups other objectives for adults use
an index of disparity based on 5 education
groups. Data for adults are for ages 25 years and
over data for adolescents are for grades 9-12.
Sources Adults - CDC, NCHS, National Health
Interview Survey Adolescents CDC, NCCDPHP,
Youth Risk Behavior Surveillance System.
24
Education groups with the best rates for PAF
objectives 2001

• Adolescents
• No moderate physical activity
• Grade 9
• No vigorous physical activity
• Grade 9
• No daily school PE
• Grade 9
• Less than 50 active PE class time
• Grade 9
• TV over 2 hours school days
• Grade 12

• Adults
• No leisure-time physical activity
• College graduate or above
• No moderate physical activity
• College graduate or above
• No vigorous physical activity
• College graduate or above
• No strength/endurance exercises
• College graduate or above
• No stretching exercises for flexibility
• At least some college

25
Changes in disparity among education groups,
selected PAF objectives Baseline to most recent
update (2001)
More disparity over time
Less disparity over time
Adults (5 groups)
Adolescents (4 grades)
-25
-20
-10
-5
0
5
10
-15
15
25
20
Percentage point change in disparity
Significantly different at Plt0.05. Index of
disparity for flexibility is based on 3 education
groups other objectives for adults use an index
of disparity based on 5 education groups. Data
for adults are for ages 25 years and over data
for adolescents are for grades 9-12. Sources
Adults - CDC, NCHS, National Health Interview
Survey (1997/98-2001) Adolescents CDC,
NCCDPHP, Youth Risk Behavior Surveillance System
(1999-2001).
26
Summary of PAF disparity measures for adults 2001

• Least disparity
• Between genders - no flexibility exercises
• Among race/ethnicity groups - no vigorous
  activity
• Among education groups - no strength and
  endurance exercises
• Most disparity
• Between genders - no leisure-time physical
  activity
• Among race/ethnicity groups - no leisure-time
  physical activity
• Among education groups - no leisure-time physical
  activity

27
Summary of PAF disparity measures for
adolescents 2001

• Least disparity
• Between genders no moderate physical activity
• Among race/ethnicity groups no daily school PE
• Among grade levels no moderate physical
  activity
• Most disparity
• Between genders no vigorous physical activity
• Among race/ethnicity groups TV watching
• Among grade levels TV watching

28
Summary of disparity for PAF objectives over
time, baseline to 2001

• Lessening disparity
• Adults gender race/ethnicity
• Adolescents race/ethnicity
• Increasing disparity
• Adults education groups
• Adolescents gender education groups

Baseline varies from 1997 to 1999 by domain and
by objective.
29
Resources

• NCHS Healthy People
• www.cdc.gov/nchs/hphome.htm
• DATA2010
• wonder.cdc.gov/data2010
• DHHS Healthy People
• www.health.gov/healthypeople

Kathleen M. Turczyn, M.P.H. Office of Analysis
and Epidemiology National Center for Health
Statistics Centers for Disease Control and
Prevention 3311 Toledo Road, Room
6323 Hyattsville, Maryland 20782 PH
301.458.4538 FX 301.458.4036 EM
kturczyn_at_cdc.gov