100
90
80
70
60
50
40
30
20
10
0
Center onSocietyand Health
Steven H. Woolf, MD, MPH*Laudan Aron, MA**Derek A. Chapman, PhD*Lisa Dubay, PhD**Emily Zimmerman, PhD*Lauren C. Snellings, MPH, CHES*Lindsey Hall, MPH*Amber D. Haley, MPH*Nikhil Holla, BA**Kristin Ayers, MPH*Christopher Lowenstein, BA**Timothy A. Waidmann, PhD**
*Center on Society and Health, Virginia Commonwealth University, Richmond, Virginia **Urban Institute, Washington, DC
Supplement 8, December 2017
SPOTLIGHT ON CHRONIC DISEASES, PART A
The
The Health of the States study, funded by the Robert Wood Johnson Foundation,
was a systematic examination of health disparities in the U.S. across the 50 states
and the District of Columbia. The study was conducted in 2014 – 2016 by the Virginia
Commonwealth University Center on Society and Health and the Urban Institute.
The goal was to take a “deep dive” into the available data on the health of the
states and the factors that shape health. The project examined how 123 potential
determinants of health, drawn from five broad domains, correlated with 39 different
health outcomes that span mortality and illness/injury across the life course.
The results were issued in a series of reports: a summary report1 released in October
2016, which was followed by a series of supplements. This report, the eighth of
nine supplements, focuses on how rates of chronic disease vary across the states.
Please refer to the first supplement 2 for details on the data sources and analytic
methods used to produce these results.
Virginia Commonwealth UniversityCenter on Society and Healthand the Urban Institute
December 2017
THE HEALTH OF THE STATES Supplement 8:
Spotlight on Chronic
Diseases, Part A
Overweight and Obesity, Diabetes, and Cardiovascular Conditions
4
in the state, and the accuracy of self-report
(versus clinical diagnosis). According
to BRFSS data, the prevalence of diabetes
in 2010 ranged 2.5 fold across the states,
from 5.4 percent in Alaska to 13.5 percent
in Alabama (Figure 1). Data on diabetes
mortality from the Centers for Disease
Control and Prevention (CDC) are more
precise because death is a discrete outcome
and based on state vital records data.
CDC data on 2013 diabetes mortality rates
also ranged 2.4 fold, from 14.1 deaths
per 100,000 in Massachusetts to 34.1 deaths
per 100,000 in West Virginia (Figure 1).
As with adult overweight/obesity,
states with the most favorable diabetes
statistics were concentrated in the Mountain
states and in New England (Figures 3–4).
Colorado and Wyoming ranked in the Top
10 on both measures — diabetes prevalence
and mortality — however Alaska had the
nation’s lowest diabetes prevalence
and Massachusetts reported the lowest
diabetes mortality rate.
States with the highest diabetes rates
were concentrated in the South and West
South Central regions. Four Southern states
Alabama, Mississippi, Tennessee, and
West Virginia — ranked in the Bottom 10
for diabetes prevalence and mortality,
as did the West South Central states of
Louisiana and Oklahoma. Alabama and
West Virginia had the nation’s highest
diabetes prevalence and mortality rates,
In 2010, the prevalence of overweight/
obesity ranged from 56.3 percent in the
District of Columbia to 69.9 percent in
Alabama (Figure 1).a The Top 10 states —
which had the nation’s lowest overweight/
obesity rates — were concentrated in the
Pacific, Mountain, and New England regions
of the country (Figure 2). The Bottom 10,
with the highest rates of overweight
and obesity, were primarily in the South
and West South Central regions.
DIABETES
Data from the Behavioral Risk Factor
Surveillance System (BRFSS) also support
estimates of the prevalence of adult
diabetes based on respondents’ answers
as to whether they have ever been told
by a doctor that they have the disease.
Statistics based on this answer can be
skewed by such factors as access to health
care, the intensity of diabetes screening
Spotlight on Chronic Diseases, Part A
a. Height and weight data
from the Behavioral
Risk Factor Surveillance
System (BRFSS) were
used to compute the
body mass index (BMI)
of adults. An overweight/
obese indicator was
computed as having a
BMI of 25.0 kg/m2 or
greater, which combined
the thresholds for over
weight (BMI = 25.0–29.9
kg/m2) and obesity
(BMI > 30.0 kg/m2).
5
Overweight/obesity
prevalence (%)
Diabetes prevalence (%)
Diabetes mortality (per
100,000)
Angina or CHD*
prevalence (%)
Heart disease mortality (per
100,000)
Stroke prevalence (%)
Stroke mortality (per
100,000)
FIGURE 1OVERWEIGHT/OBESITY, DIABETES, AND CARDIOVASCULAR CONDITIONS (PER 100,000), BY STATE
DC 56.3 AK 5.4 MA 14.1 HI 2.3 MN 166.5 CT 1.7 NY 26.3HI 57.2 CO 6.1 WY 14.2 DC 2.6 CO 172.0 CO 1.8 NH 27.6
CO 57.6 UT 6.6 CT 14.8 AK 2.6 MA 182.6 MN 1.9 MA 27.7UT 57.7 MN 6.8 NV 14.8 UT 2.8 HI 185.3 WY 1.9 RI 27.7VT 58.4 VT 6.9 CO 15.0 CO 3.1 AZ 186.7 WI 2.0 CT 28.3
MA 60.1 SD 7.0 HI 15.5 WA 3.4 AK 186.9 AK 2.0 AZ 28.6NV 60.2 MT 7.2 VT 17.4 GA 3.5 WA 188.5 MA 2.1 NM 30.0CT 60.6 WI 7.2 NY 17.8 CA 3.6 NH 188.8 SD 2.1 DC 30.1
NM 60.7 WY 7.3 DC 17.9 CT 3.6 OR 189.6 NH 2.2 FL 30.6OR 60.9 OR 7.3 VA 18.4 ID 3.6 NM 190.6 UT 2.2 VT 31.7VA 61.3 CT 7.4 WI 18.5 OR 3.6 CT 192.3 VT 2.2 CO 32.0MT 61.3 ND 7.6 NH 18.7 MN 3.6 FL 195.2 NY 2.2 MN 32.0NY 61.5 MA 7.6 MN 18.8 MT 3.7 ID 196.4 ID 2.2 ND 32.4NJ 61.5 IA 7.7 RI 18.9 MD 3.7 VT 196.9 WA 2.3 NJ 32.4CA 61.6 NE 7.8 IA 19.0 VT 3.7 ME 198.4 CA 2.3 NV 33.3WA 61.8 WA 7.8 MD 19.1 WI 3.8 UT 199.3 ND 2.4 ME 33.4
ID 62.9 RI 8.1 FL 19.2 NJ 3.9 ND 200.1 NE 2.4 IA 33.9MN 63.1 ID 8.1 DE 19.4 IL 3.9 NE 200.9 NJ 2.4 HI 34.8NH 63.1 NH 8.1 NJ 19.4 WY 3.9 WY 201.3 RI 2.4 CA 34.9
IL 63.2 DC 8.4 IL 19.7 NV 4.0 SD 202.3 MD 2.5 WY 35.1RI 63.5 HI 8.5 MT 19.7 NE 4.0 RI 202.7 OR 2.5 ID 35.4
WI 63.6 NM 8.6 KS 19.8 IA 4.0 CA 205.0 HI 2.5 WA 35.5ME 63.7 KS 8.6 AK 20.1 NM 4.0 MT 206.2 KS 2.6 WI 36.0WY 63.8 NV 8.7 ME 20.4 VA 4.1 WI 209.7 NM 2.6 MD 36.1DE 64.0 CA 8.9 MO 20.5 AZ 4.1 VA 209.9 VA 2.7 MI 36.3KS 64.6 ME 8.9 CA 20.6 MA 4.1 DE 217.1 IL 2.7 NE 36.4
ND 64.8 VA 8.9 WA 21.3 KS 4.1 NJ 217.3 GA 2.8 IL 36.7AZ 64.8 DE 8.9 TX 21.6 RI 4.2 KS 217.6 MT 2.8 DE 37.0NE 64.9 IL 8.9 NE 21.7 OH 4.3 IA 218.6 ME 2.8 PA 37.2FL 65.0 NY 9.0 NC 21.8 ND 4.4 IL 222.7 IA 2.8 OR 37.3
NC 65.3 AZ 9.2 ND 22.3 NY 4.4 MD 223.1 TX 2.8 MT 37.6OH 65.7 NJ 9.4 SC 22.5 MS 4.4 NC 223.1 MI 2.9 SD 38.0SD 65.7 MD 9.5 PA 22.6 NH 4.4 TX 225.0 IN 2.9 KS 38.1PA 65.7 MO 9.7 SD 22.7 TX 4.4 NY 227.2 DE 3.0 UT 38.2
GA 65.8 AR 9.8 GA 23.0 DE 4.5 PA 230.3 OH 3.0 VA 38.6MO 65.8 GA 9.9 OR 23.4 MO 4.6 GA 240.1 NC 3.1 OH 39.9AK 65.9 TX 9.9 AZ 23.5 IN 4.6 SC 243.1 NV 3.1 TX 40.2
MD 66.0 NC 10.0 ID 23.7 SC 4.6 NV 244.0 AZ 3.2 MO 40.6IA 66.2 IN 10.0 MI 23.8 SD 4.6 IN 244.7 PA 3.4 AK 40.7IN 66.5 KY 10.3 KY 24.1 NC 4.6 OH 245.6 LA 3.4 IN 40.7LA 66.5 MI 10.3 AR 24.2 PA 4.7 MO 249.3 DC 3.4 WV 40.7TX 66.6 OH 10.4 AL 24.3 TN 4.8 MI 254.1 TN 3.5 GA 41.4MI 66.8 LA 10.5 TN 24.8 AR 5.1 WV 257.7 WV 3.5 KY 41.7AR 67.1 PA 10.5 UT 25.3 ME 5.2 KY 261.1 KY 3.5 NC 42.4OK 67.2 FL 10.6 OH 25.4 MI 5.3 DC 261.3 FL 3.5 LA 44.0SC 67.4 OK 10.7 IN 26.3 LA 5.3 TN 265.8 AR 3.6 TN 44.4KY 67.5 SC 11.0 LA 26.9 OK 5.4 LA 275.5 SC 3.7 OK 44.5TN 67.8 WV 11.8 NM 27.6 FL 5.5 AR 278.3 MO 3.9 MS 47.2WV 67.9 TN 11.9 OK 29.9 AL 5.6 OK 289.5 MS 4.1 AR 47.6MS 68.8 MS 12.6 MS 32.9 KY 5.8 AL 297.1 OK 4.2 SC 47.6AL 69.9 AL 13.5 WV 34.1 WV 6.0 MS 308.3 AL 4.7 AL 48.1
6
CARDIOVASCULAR DISEASE
Cardiovascular disease was measured
in terms of prevalence and mortality.
The BRFSS asks respondents about heart
disease — whether they have ever been
told by a doctor that they had angina
or [coronary] heart disease — as well as
cerebrovascular diseases (e.g., stroke).
In contrast, CDC mortality data measure
the death rate from cardiovascular
diseases, a category that encompasses
heart disease and other vascular conditions,
such as cerebrovascular disease (e.g.,
stroke). Cerebrovascular disease, a form
of cardiovascular disease that damages
the brain and is itself a leading cause
of death, was also measured in terms of
prevalence and mortality.
respectively (Figure 1). In the Midwest,
Ohio also ranked in the Bottom 10 on
both measures.
Utah, which ranked in the Top 10
for low obesity rates (see above), ranked
in the Bottom 10 for diabetes mortality.
This is all the more surprising given
that Utah ranked in the Top 10 for its low
diabetes prevalence; it also ranked in
the Top 10 for 20 other health outcomes
examined in this report, a total matched
only by Massachusetts. The District of
Columbia, which struggles with other
health disadvan tages, achieved the Top 10
for both obesity and diabetes mortality.
Several Southern states had distinctly
higher diabetes prevalence and mortality
rates than other states in the Bottom 10
(Figure 1).
FIGURE 2PREVALENCE OF ADULT OVERWEIGHT/OBESITY (PERCENT), BY STATE (2010)
CENSUS REGIONS
STATE (Rank)
Pacifi cM
ountainW
. So. CentralW
. No. CentralE. No. CentralNew EnglandM
iddle AtlanticSouth
TOP 10 STATES: LOWEST RATES OF ADULT OVERWEIGHT / OBESITY
Washington, D.C. (1) xHawaii (2) xColorado (3) xUtah (4) xVermont (5) xMassachusetts (6) xNevada (7) xConnecticut (8) xNew Mexico (9) x Oregon (10) xBOTTOM 10 STATES: HIGHEST RATES OF
ADULT OVERWEIGHT/OBESITYAlabama (51) xMississippi (50) xWest Virginia (49) xTennessee (48) xKentucky (47) xSouth Carolina (46) xOklahoma (45) xArkansas (44) xMichigan (43) xTexas (42) x
Adult overweight/ obes ity, 2010
56.3 – 61.3
61.4 – 63.5
63.6 – 65.3
65.4 – 66.5
66.6 – 69.9
7
Alabama (Figure 1). States with the lowest
prevalence of angina and coronary heart
disease and the lowest cardiovascular
mortality rates were primarily in the Pacific,
Mountain, and New England regions
(Figures 5–6). States in these regions also
had among the lowest rates of strokes and
cerebrovascular mortality (Figures 7–8).b
Pacific RegionAlaska and Washington ranked in
the Top 10 for their low prevalence of heart
disease and cardiovascular mortality.
Alaska ranked for its low prevalence of
strokes. Hawaii ranked in the Top 10 for
its low cardiovascular and cerebrovascular
mortality rates; it also had the nation’s
lowest prevalence of heart disease (Figure 1).
b. We also note the
following findings
elsewhere in the country:
The District of Columbia
ranked in the Top 10
on two measures —
low prevalence of heart
disease and low cerebro
vascular mortality.
Minnesota had the
lowest cardiovascular
mortality rate in the
nation, whereas New
York had the lowest
cerebrovascular mortality
rate in the nation.
In the 2010 BRFSS survey, the prevalence
of heart disease ranged almost three-fold
across the states, from 2.3 percent in
Hawaii to 6.0 percent in West Virginia,
and the prevalence of stroke varied from
1.7 percent in Connecticut to 4.7 percent
in Alabama (Figure 1). As with diabetes,
comparing prevalence rates is problematic
because these measures may in fact be
reflecting differences across states in
access to health care and disease screening
rather than true differences in prevalence.
Death rates in 2013 from cardiovascular
disease ranged from 166.5 deaths per
100,000 in Minnesota to 308.3 deaths per
100,000 in Mississippi (Figure 1); cardio-
vascular deaths due to cerebrovascular
disease ranged from 26.3 per 100,000 in
New York to 48.1 deaths per 100,000 in
FIGURE 3DIABETES PREVALENCE (PERCENT), BY STATE (2010)
CENSUS REGIONS
STATE (Rank)
Pacifi cM
ountainW
. So. CentralW
. No. CentralE. No. CentralNew EnglandM
iddle AtlanticSouth
TOP 10 STATES: LOWEST DIABETES PREVALENCE
Alaska (1) x Colorado (2) xUtah (3) xMinnesota (4) xVermont (5) xSouth Dakota (6) xMontana (7) xWisconsin (8) xWyoming (9) x Oregon (10) x
BOTTOM 10 STATES: HIGHEST DIABETES PREVALENCE
Alabama (51) xMississippi (50) xTennessee (49) xWest Virginia (48) xSouth Carolina (47) xOklahoma (46) xFlorida (45) xPennsylvania (44) xLouisiana (43) xOhio (42) x
Diabetes P revalence 2010
5.4 – 7.4
7.5 – 8.5
8.6 – 9.2
9.3 – 10.3
10.4 – 13.5
8
FIGURE 4DIABETES MORTALITY (PER 100,000), BY STATE (2013)
CENSUS REGIONS
STATE (Rank)
Pacifi cM
ountainW
. So. CentralW
. No. CentralE. No. CentralNew EnglandM
iddle AtlanticSouth
TOP 10 STATES: LOWEST DIABETES MORTALIT Y
Massachusetts (1) x Wyoming (2) xNevada (3) xConnecticut (3) xColorado (5) xHawaii (6) xVermont (7) xNew York (8) xWashington, D.C. (9) xVirginia (10) x
BOTTOM 10 STATES: HIGHEST DIABETES MORTALIT Y
West Virginia (51) xMississippi (50) xOklahoma (49) xNew Mexico (48) xLouisiana (47) xIndiana (46) xOhio (45) xUtah (44) xTennessee (43) xAlabama (42) x
Diabetes morta lity per 100,000 2013
14.1 – 18.5
18.6 – 19.7
19.8 – 22.3
22.4 – 24.2
24.3 – 34.1
FIGURE 5PREVALENCE OF ANGINA/CORONARY HEART DISEASE (PERCENT), BY STATE (2010)
CENSUS REGIONS
STATE (Rank)
Pacifi cM
ountainW
. So. CentralW
. No. CentralE. No. CentralNew EnglandM
iddle AtlanticSouth
TOP 10 STATES: LOWEST PREVALENCE OF ANGINA OR CORONARY HEART DISEASE
Hawaii (1) xWashington, D.C. (2) xAlaska (3) xUtah (4) xColorado (5) xWashington (6) xGeorgia (7) xCalifornia (8) xConnecticut (8) x Idaho (10) xBOTTOM 10 STATES: HIGHEST PREVALENCE OF ANGINA OR CORONARY HEART DISEASE
West Virginia (51) xKentucky (50) xAlabama (49) xFlorida (48) xOklahoma (47) xLouisiana (46) xMichigan (45) xMaine (44) xArkansas (43) xTennessee (42) xSee Supplement 1: The Health of the States: Spotlight on
Methods for our protocol for handling tied rankings.
Heart Dis eas e prevalence 2010
2.3 – 3.6
3.7 – 4.0
4.1 – 4.4
4.5 – 4.7
4.8 – 6.0
9
FIGURE 6CARDIOVASCULAR DISEASE MORTALITY (PER 100,000), BY STATE (2013)
CENSUS REGIONS
STATE (Rank)
Pacifi cM
ountainW
. So. CentralW
. No. CentralE. No. CentralNew EnglandM
iddle AtlanticSouth
TOP 10 STATES: LOWEST CARDIOVASCUL AR DISEASE MORTALIT Y
Minnesota (1) xColorado (2) xMassachusetts (3) xHawaii (4) xArizona (5) xAlaska (6) xWashington (7) xNew Hampshire (8) xOregon (9) x New Mexico (10) x
BOTTOM 10 STATES: HIGHEST CARDIOVASCUL AR DISEASE MORTALIT Y
Mississippi (51) xAlabama (50) xOklahoma (49) xArkansas (48) xLouisiana (47) xTennessee (46) xWashington, D.C. (45) xKentucky (44) xWest Virginia (43) xMichigan (42) x
C ardiovas cular mortality per 100,000 2013
166.5 – 192.3
192.4 – 202.7
202.8 – 223.1
223.2 – 254.1
254.2 – 308.3
FIGURE 7PREVALENCE OF STROKE (PERCENT), BY STATE (2010)
CENSUS REGIONS
STATE (Rank)
Pacifi c
Mountain
W. So. Central
W. No. Central
E. No. Central
New England
Middle Atlantic
South
TOP 10 STATES: LOWEST STROKE PREVALENCE
Connecticut (1) xColorado (2) xMinnesota (3) xWyoming (4) xWisconsin (5) xAlaska (6) xMassachusetts (7) xSouth Dakota (8) xNew Hampshire (9) x Utah (10) xVermont (10) x
BOTTOM 10 STATES: HIGHEST STROKE PREVALENCE
Alabama (51) xOklahoma (50) xMississippi (49) xMissouri (48) xSouth Carolina (47) xArkansas (46) xFlorida (45) xKentucky (44) xWest Virginia (43) xTennessee (42) xPercentage of persons age 18 years and older who
reported ever having had a stroke. Top 10 for this
outcome includes 11 states, Bottom 10 includes 10 states.
See Supplement 1: The Health of the States: Spotlight on
Methods for our protocol for handling tied rankings.
S troke prevelance 2010
1.7 – 2.2
2.3 – 2.5
2.6 – 2.8
2.9 – 3.4
3.5 – 4.7
10
and cerebrovascular mortality and stroke
prevalence).
Sixteen states, dominated by Southern and
West South Central states in the “stroke
belt,” ranked in the Bottom 10 for at least
one of the four measures: the prevalence
of heart disease, cardiovascular mortality,
stroke prevalence, or cerebrovascular
mortality. Alabama, Arkansas, Oklahoma,
Tennessee, and Kentucky ranked in the
Bottom 10 on all four measures.c Figure 1
also reveals the following:
• Alabama had the nation’s highest
rates of stroke prevalence and cere-
brovascular mortality, and the second
highest cardiovascular mortality.
c. Michigan, an East North
Central state, ranked
in the Bottom 10 for
its prevalence of heart
disease and cardio
vascular death rate.
Mountain Region Colorado ranked in the Top 10 for three
measures (the low prevalence of heart
disease and strokes and low cardiovascular
mortality), whereas Utah ranked in the
Top 10 for two measures (low prevalence
of heart disease and strokes), as did
Arizona and New Mexico (low cardiovas-
cular and cerebrovascular mortality).
New England Region New England states performed better
around cerebrovascular disease
than cardiovascular disease overall.
Connecticut had the lowest prevalence
of strokes in the nation. New Hampshire
and Massachusetts ranked in the Top
10 on three measures (cardiovascular
FIGURE 8CEREBROVASCULAR MORTALITY (PER 100,000), BY STATE (2013)
CENSUS REGIONS
STATE (Rank)
Pacifi cM
ountainW
. So. CentralW
. No. CentralE. No. CentralNew EnglandM
iddle AtlanticSouth
TOP 10 STATES: LOWEST CEREBROVASCUL AR MORTALIT Y
New York (1) xNew Hampshire (2) xRhode Island (3) xMassachusetts (3) xConnecticut (5) xArizona (6) xNew Mexico (7) xWashington, D.C. (8) xFlorida (9) xVermont (10) x
BOTTOM 10 STATES: HIGHEST CEREBROVASCUL AR MORTALIT Y
Alabama (51) xSouth Carolina (49) xArkansas (49) xMississippi (48) xOklahoma (47) xTennessee (46) xLouisiana (45) xNorth Carolina (44) xKentucky (43) xGeorgia (42) xSee Supplement 1: The Health of the States: Spotlight on
Methods for our protocol for handling tied rankings.
C erebrovascular morta lity per 100,000 2013
26.3 - 32.0
32.1 - 35.5
35.6 - 38.0
38.1 - 41.4
41.5 - 48.1
11
• Mississippi had the nation’s
highest cardiovascular death rate.
It also ranked in the Bottom 10
on stroke prevalence and cerebro-
vascular mortality.
• West Virginia had the nation’s
highest prevalence of heart disease.
It also ranked in the Bottom 10
on stroke prevalence and cardio-
vascular mortality.
• South Carolina ranked in the Bottom
10 on stroke prevalence and cerebro-
vascular mortality.
• Florida ranked in the Bottom 10 only
for the prevalence of heart disease
and stroke but, as noted above,
achieved a Top 10 ranking for low
cerebro vascular morality.
Some geographic patterns were noteworthy
for their discrepancies. For example, some
Pacific states with Top 10 standing for
cardio vascular disease and heart disease
prevalence did not always report the lowest
rates for cerebrovascular disease. As just
noted, Florida reported a high prevalence
of strokes but low stroke mortality. The
neighboring state of Georgia reported a low
prevalence of heart disease but ranked in
the Bottom 10 for its high cerebrovascular
mortality. The District of Columbia —
which ranked in the Top 10 for its low
prevalence of obesity and heart disease,
and its low mortality rates for diabetes and
FIGURE 9
WHAT CORRELATES WITH ADULT OVERWEIGHT/OBESITY? THE CORRELATION COEFFICIENTS (rs)*
HEALTH BEHAVIORS
Physical inactivity (adult) 0.83 Any breastfeeding -0.70
Soda intake (youth) 0.70 Bicycle helmet use (youth) -0.65
Current smokers 0.68 Physical activity (children) -0.54
Sexual activity before age 18 0.55 Fruit intake (youth) -0.53
PHYSICAL AND SOCIAL ENVIRONMENT
Commuting by motor vehicle 0.78 Neighborhood resources for children -0.74
Smokers in household (child present) 0.73 Neighborhoods that
are walkable -0.64
Proximity to parks -0.58
SOCIAL AND ECONOMIC FACTORS
Severe housing disrepair 0.65 Bachelor’s degree/higher -0.68
Poverty (adults) 0.53 Median household income -0.64
Higher educated household head -0.53
HEALTH SYSTEM
Avoidable hospitalization 0.73 Annual dental visit (adult) -0.54
Primary care shortage 0.63
PUBLIC POLICIES AND SPENDING
Unemployment benefits ÷ pop. <100% FPL -0.55
State income support ÷ pop. <100% FPL -0.54
Mass transit spending per capita -0.51
*Correlation coefficients range from zero to 1.0 and measure how strongly one variable correlates with another. Factors on the right (negative coefficients) are inversely related (e.g., one goes up when the other goes down).
High correlations were also noted for other measures of Physical and Social Environment: Neighborhoods that are walkable (rs= -0.58), Commuting by public transit (-0.56); Social and Economic Factors: Poverty (children) (0.59), Residents in very concentrated (>40%) poverty (0.52); and Public Policies and Spending: State/Federal income support ÷ pop. <100% FPL (-0.57); Unemployment benefits ÷ pop. <200% FPL (-0.55), State income support ÷ pop. <200% FPL (-0.54); Public welfare workers ÷ pop. <200% FPL (-0.53).
See Supplement 1: The Health of the States: Spotlight on methods for definitions of terms, data sources, and methods for calculating the correlation coefficients.
x
x
x
xxx
12
cerebrovascular disease, ranked in the
Bottom 10 for cardiovascular mortality.
Although cardiovascular and cerebro-
vascular conditions share many common
risk factors (e.g., smoking, hypertension,
obesity) and would be expected to share
similar geographic patterns, other factors
could explain discrepancies. For example,
geographic differences in cerebrovascular
mortality may be due to the speed with
which stroke victims receive treatment.
Whether these discrepancies reflect a
difference in treatment success rates or in
access to care is unclear.
What correlates the most
with obesity, diabetes, and
cardiovascular conditions?
As reported widely in the literature,
the prevalence of overweight/obesity,
diabetes, and cardiovascular conditions
correlated with unhealthy behaviors
(see Figures 9–15). For example, states
that ranked highly on physical inactivity
had higher rates of overweight/obesity,
diabetes prevalence and mortality, angina/
coronary heart disease, cardiovascular
mortality, strokes, and cerebrovascular
mortality. The prevalence of adult physical
inactivity was 29.4 percent in Bottom 10
states for adult overweight/obesity, com-
pared with 19.7 percent in Top 10 states.
Cigarette smoking also correlated with the
prevalence and mortality rates for these
diseases. For example, the percentage of
FIGURE 10
WHAT CORRELATES WITH DIABETES PREVALENCE? THE CORRELATION COEFFICIENTS (rs)*
HEALTH BEHAVIORS
Physical inactivity (adult) 0.78 Exclusive breastfeeding -0.76
Screen time (youth) 0.65 Bicycle helmet use (youth) -0.66
Sexual activity before age 18 0.58 Fruit intake (youth) -0.59
Soda intake (youth) 0.55 Physical activity (children) -0.59
Current nonsmokers -0.57
PHYSICAL AND SOCIAL ENVIRONMENT
Commuting by motor vehicle 0.63 Safe schools (parent report) -0.68
Indoor smoking (child present) 0.57 Neighborhood resources for children -0.67
Dating violence (youth) 0.53 Residents in walkable neighborhoods -0.56
Proximity to parks -0.54
SOCIAL AND ECONOMIC FACTORS
Poverty (children) 0.69 Employment -0.77
Poverty (adults) 0.67 Higher educated household head -0.59
Residents in concentrated (>20%) poverty 0.64 Median household income -0.57
Single-parent households 0.62 Proficient in math (grade 8) -0.53Poor living in concentrated (>20%) poverty 0.61 Bachelor’s degree/higher -0.50
Food insecurity (households) 0.54
Income inequality 0.53
HEALTH SYSTEM
Avoidable hospitalization 0.73 Private insurance -0.61
Could not afford doctor 0.65 Annual dental visit (adult) -0.53
Rehospitalization 0.59 Electronic health record system -0.51
PUBLIC POLICIES AND SPENDINGState income support ÷ pop. <100% FPL -0.53
Libraries (per capita) -0.50
*Correlation coefficients range from zero to 1.0 and measure how strongly one variable correlates with another. Factors on the right (negative coefficients) are inversely related (e.g., one goes up when the other goes down).
High correlations were also noted for Health Behaviors: Any breastfeeding (rs= -0.69); Physical and Social Environment: Smoke-free homes (-0.54), Commuting by walking/cycling (-0.74), Smoking in home (nonsmokers present) (0.54), Neighborhoods that are walkable (-0.52); Social and Economic Factors: Residents in very concentrated (>40%) poverty (0.56), Children with employed parents (-0.55), Poverty (supplemental def.) (0.51); Health Systems: Rehospitalization (heart failure) (0.60), Rehospitalization (pneumonia) (0.58), Rehospitalization (heart attack) (0.52). Diabetes prevalence correlated with spending on libraries, calculated per capita (rs= -0.50); and Public Policies and Spending: Libraries ÷ pop. <100% FPL (-0.63), Libraries ÷ pop. <200% FPL (-0.59), Social services ÷ pop. <100% FPL (-0.53).
See Supplement 1: The Health of the States: Spotlight on methods for definitions of terms, data sources, and methods for calculating the correlation coefficients.
x
x
xx
xx
13
FIGURE 12
WHAT CORRELATES WITH THE PREVALENCE OF ANGINA/CORONARY HEART DISEASE?THE CORRELATION COEFFICIENTS (rs)*
HEALTH BEHAVIORS
Physical inactivity (adult) 0.74 Any breastfeeding -0.61
Soda intake (youth) 0.64 Sexual abstinence before age 18 -0.52
Current smokers 0.63
PHYSICAL AND SOCIAL ENVIRONMENT
Commuting by motor vehicle 0.72 Neighborhood
resources for children -0.68
Distance to parks 0.64 Smoke-free homes -0.66
Residents in walkable neighborhoods -0.53
SOCIAL AND ECONOMIC FACTORS
Poverty (adults) 0.54 Median household income -0.67
Severe housing disrepair 0.52 Bachelor’s degree/higher -0.54
Employment -0.51
HEALTH SYSTEM
Avoidable hospitalization 0.75
PUBLIC POLICIES AND SPENDING
State income support ÷ pop. <100% FPL -0.54
Unemployment benefits ÷ pop. <100% FPL -0.52
*Correlation coefficients range from zero to 1.0 and measure how strongly one variable correlates with another. Factors on the right (negative coefficients) are inversely related (e.g., one goes up when the other goes down.
High correlations were also noted for other measures of Health Behaviors: Ever smokers (rs= 0.61); Physical and Social Environment: Smokers in household (child present) (0.65), Indoor smoking (child present) (0.62), Commuting by walking/cycling (-0.57), Indoor smoking (nonsmokers present) (0.54), Commuting by public transit (-0.54); and Public Policies and Spending: State income support ÷ pop. <200% FPL (-0.51).
See Supplement 1: The Health of the States: Spotlight on methods for definitions of terms, data sources, and methods for calculating the correlation coefficients.
x
x
x
x
xx
FIGURE 11
WHAT CORRELATES WITH DIABETES MORTALITY? THE CORRELATION COEFFICIENTS (rs)*
HEALTH BEHAVIORS
Physical inactivity (adult) 0.54
PHYSICAL AND SOCIAL ENVIRONMENT
Commuting by motor vehicle 0.63 Neighborhood
resources for children -0.62
Childhood trauma 0.60 Residents in walkable neighborhoods -0.60
Smokers in household (child present) 0.51
SOCIAL AND ECONOMIC FACTORS
Poverty (adults) 0.69 Median household income -0.74
Residents in concentrated (>20%) poverty 0.62 Bachelor’s degree/higher -0.73
Severe housing disrepair 0.59 Employment -0.60
Food insecurity (households) 0.51 Higher educated
household head -0.56
Proficient in reading (grade 4) -0.51
HEALTH SYSTEM
Primary care shortage 0.68 Annual dental visit (adult) -0.58
Could not afford doctor 0.62 Private insurance -0.54
PUBLIC POLICIES AND SPENDING
Unemployment benefits ÷ pop. <100% -0.62
State income support ÷ pop. <100% FPL -0.59
Public welfare workers ÷ pop. <100% FPL -0.55
*Correlation coefficients range from zero to 1.0 and measure how strongly one variable correlates with another. Factors on the right (negative coefficients) are inversely related (e.g., one goes up when the other goes down).
High correlations were also noted for other measures of Physical and Social Environment: Neighborhoods that are walkable (rs= -0.58), Commuting by public transit (-0.56); Social and Economic Factors: Poverty (children) (0.59), Residents in very concentrated (>40%) poverty (0.52); and Public Policies and Spending: State/Federal income support ÷ pop. <100% FPL (-0.57); Unemployment benefits ÷ pop. <200% FPL (-0.55), State income support ÷ pop. <200% FPL (-0.54); Public welfare workers ÷ pop. <200% FPL (-0.53).
See Supplement 1: The Health of the States: Spotlight on methods for definitions of terms, data sources, and methods for calculating the correlation coefficients.
x
x
x
xxx
14
adults who were current smokers was
21.7 percent in Bottom 10 states for high
rates of angina/coronary heart disease,
compared with 14.9 percent in Top 10
states with low heart disease rates.
But states with higher disease rates
were also places where other unhealthy
behaviors were more common. For
example, in these states, women were
less likely to breastfeed, and children
and teens had unhealthier diets (e.g.,
less fruit, more soda), were less likely
to be physically active, and became
sexually active at an earlier age. These
associations do not necessarily reflect
causal relationships but rather a pattern
of co-occurrence, where conditions “go
together” at the state level. States where
people often engage in a behavior that
causes one disease may also rank highly
on behaviors that cause other diseases
or injuries.
States where unhealthy behaviors
were more prevalent were also more
likely to have unhealthier physical
environments, which also correlated
highly with obesity, diabetes, and
cardiovascular diseases.3 For example,
as seen in Figures 9–15, states where
more residents commuted to work by
motor vehicle (rather than by walking,
cycling, or public transportation) had a
higher prevalence of overweight/obesity,
diabetes, angina/coronary heart disease,
and strokes, as well as higher diabetes,
cardiovascular, and cerebrovascular
mortality. Conversely, these conditions
FIGURE 13
WHAT CORRELATES WITH CARDIOVASCULAR MORTALITY?THE CORRELATION COEFFICIENTS (rs)*
HEALTH BEHAVIORS
Physical inactivity (adult) 0.81 Any breastfeeding -0.74
Current smokers 0.69 Bicycle helmet use (youth) -0.66
Soda intake (youth) 0.58 Fruit intake (youth) -0.60
Sexual abstinence before age 18 -0.54
PHYSICAL AND SOCIAL ENVIRONMENT
Indoor smoking (child present) 0.77 Neighborhood resources for children -0.52
Commuting by motor vehicle 0.62 Safe schools (parent report) -0.51
Air pollution 0.50
SOCIAL AND ECONOMIC FACTORS
Poverty (adults) 0.64 Proficient in math (grade 8) -0.55
Residents in concentrated (>20%) poverty 0.62 Median household income -0.55
Poor living in concentrated (>20%) poverty 0.58 Bachelor’s degree/higher -0.55
Single-parent households 0.51 Higher educated household head -0.55
Severe housing disrepair 0.50 Employment -0.50
HEALTH SYSTEM
Avoidable hospitalization 0.75 Can afford doctor -0.52
Primary care shortage 0.59 Electronic health record system -0.51
Rehospitalization (heart failure) 0.59
*Correlation coefficients range from zero to 1.0 and measure how strongly one variable correlates with another. Factors on the right (negative coefficients) are inversely related (e.g., one goes up when the other goes down).
High correlations were also noted for other measures of Health Behaviors: Exclusive breastfeeding (rs= -0.69); Physical and Social Environment: Indoor smoking (nonsmokers present) (0.63), Smokers in household (child present) (0.62), Commuting by walking/cycling (-0.60); Social and Economic Factors: Poverty (children) (0.63), Residents in very concentrated (>40%) poverty (0.59), Proficient in math (grade 4) (-0.52); and Health Systems: Rehospitalization (pneumonia) (0.54), Rehospitalization (0.52).
See Supplement 1: The Health of the States: Spotlight on methods for definitions of terms, data sources, and methods for calculating the correlation coefficients.
x
x
x
15
FIGURE 14
WHAT CORRELATES WITH THE PREVALENCE OF STROKES?THE CORRELATION COEFFICIENTS (rs)*
HEALTH BEHAVIORS
Physical inactivity (adult) 0.71 Exclusive breastfeeding -0.67
Sexual activity before age 18 0.69 Bicycle helmet use (youth) -0.67
Current smokers 0.65 Fruit intake (youth) -0.62
Soda intake (youth) 0.58 Physical activity (children) -0.56
Birth control (youth) -0.50
PHYSICAL AND SOCIAL ENVIRONMENT
Commuting by motor vehicle 0.65 Safe schools (parent report) -0.65
Smokers in household (child present) 0.60 Neighborhood resources
for children -0.64
Childhood trauma 0.60 Residents in walkable neighborhoods -0.60
Proximity to parks -0.50
SOCIAL AND ECONOMIC FACTORS
Poverty (children) 0.76 Employment -0.72
Poverty (adults) 0.74 Median household income -0.69Residents in concentrated (>20%) poverty 0.69 Higher educated
household head -0.68
Single-parent households 0.59 Proficient in math (grade8) -0.63Poor living in concentrated (>20%) poverty 0.59 Bachelor’s degree/higher -0.57
Food insecurity (households) 0.58
Adults in prison 0.51
HEALTH SYSTEM
Avoidable hospitalization 0.60 Private insurance -0.66
Could not afford doctor 0.59 Annual dental visit (adult) -0.64
PUBLIC POLICIES AND SPENDINGState income support ÷ pop. <100% FPL -0.61Federal public assistance ÷ pop. <100% FPL -0.53Public welfare workers ÷ pop. <100% FPL -0.53
*Correlation coefficients range from zero to 1.0 and measure how strongly one variable correlates with another. Factors on the right (negative coefficients) are inversely related (e.g., one goes up when the other goes down).
High correlations were also noted for other measures of Health Behaviors: Any breastfeeding (rs= -0.64); Physical and Social Environment: Commuting by walking/cycling (-0.64), Indoor smoking (child present) (0.57), Neighborhoods that are walkable (-0.57), Smoke-free homes (-0.56), Indoor smoking (nonsmokers present) (0.53); Social and Economic Factors: Proficient in reading (grade 8) (-0.59), Proficient in math (grade 4) (-0.56), Children with employed parents (-0.55); Health Systems: Public insurance (0.53); and Public Policies and Spending: State income support ÷ pop. <200% FPL (-0.57); State/Federal income support ÷ pop. <100% FPL (-0.52); Federal public assistance ÷ pop. <200% FPL (-0.51). Prevalence correlated with spending on libraries, calculated per capita (-0.48).
x
x
xx
xxx
FIGURE 15
WHAT CORRELATES WITH CEREBROVASCULAR MORTALITY?THE CORRELATION COEFFICIENTS (rs)*
HEALTH BEHAVIORS
Current smokers 0.67 Bicycle helmet use (youth) -0.56
Physical inactivity (adult) 0.67 Any breastfeeding -0.53
Soda intake (youth) 0.61
Sexual activity before age 18 0.60
PHYSICAL AND SOCIAL ENVIRONMENT
Commuting by motor vehicle 0.65 Neighborhoods that
are walkable -0.67
Smokers in household (child present) 0.63 Neighborhood resources
for children -0.66
SOCIAL AND ECONOMIC FACTORS
Severe housing disrepair 0.58 Bachelor’s degree/higher -0.65
Adults in prison 0.56 Median household income -0.60
Food insecurity (households) 0.51 Higher educated
household head -0.56
Poverty (adults) 0.50
HEALTH SYSTEM
Primary care shortage 0.74 Annual dental visit (adult) -0.61
Avoidable hospitalization 0.52 Could afford doctor -0.60
PUBLIC POLICIES AND SPENDING
Tobacco taxes -0.64
Medicaid eligibility limits -0.59
State income support ÷ pop. <100% FPL -0.56
Unemployment benefits ÷ pop. <100% -0.50
*Correlation coefficients range from zero to 1.0 and measure how strongly one variable correlates with another. Factors on the right (negative coefficients) are inversely related (e.g., one goes up when the other goes down).
High correlations were also noted for other measures of Physical and Social Environment: Residents in walkable neighborhoods (rs= -0.68), Commuting by walking/cycling (-0.58), Indoor smoking (child present) (0.53), Commuting by public transit (-0.50); and Public Policies and Spending: Medicaid eligibility (other) (-0.58), State/Federal income support ÷ pop. <100% FPL (-0.50).
See Supplement 1: The Health of the States: Spotlight on methods for definitions of terms, data sources, and methods for calculating the correlation coefficients.
xx
x
xxxx
16
In Bottom 10 states (with high stroke rates),
17.2 percent of the population could not
afford their doctor, compared with 10.2
percent in Top 10 states with the lowest
stroke rates. States with high rates of
avoidable hospitalizations — suggesting
inadequate primary care — ranked higher
on overweight/obesity, diabetes, angina/
coronary heart disease, cardiovascular
mortality, strokes, and cerebrovascular
mortality (see Figure 17). Rates for hospital
readmissions within 30 days — also
suggesting inadequate outpatient disease
management — also correlated with
diabetes prevalence and cardiovascular
mortality. There were fewer annual dental
visits in these states, another marker
for inadequate access to health care.
Diabetes and strokes were more common
in states where residents lacked private
health insurance and where more patients
reported being unable to afford their
doctor (Figures 10 and 13).
Socioeconomic conditions are
a powerful explanatory factor for these
correlations.3 State rankings for obesity,
diabetes, and cardiovascular conditions
correlated very highly with a variety
of measures of employment, income, and
education (Figures 9–15). For example,
in Top 10 states for low diabetes mortality,
more than one out of four adults (26.0
percent) had a Bachelor’s degree or higher,
compared with 16.6 percent in Bottom
10 states. In states where the prevalence
were less common in states where residents
had less exposure to secondhand smoke
at home, a healthier built environment, and
more walkable neighborhoods.
For example, in Top 10 states (with
low rates of adult overweight/obesity),
walkable neighborhoods were more than
five times more prevalent than in Bottom
10 states, commuting by public transpor-
tation was almost 10 times more common,
and access to parks was more than twice as
great (Figure 16). As reported elsewhere,4,5
air pollution may also matter; state rank-
ings for fine particulate matter correlated
with cardiovascular mortality rates.
The indicators available in our data
suggest that the social environment
may also be unhealthier in states where
cardiovascular diseases were more common.
School safety as perceived by parents
was poorer in states with higher rates
of diabetes (Figure 10), cardiovascular
mortality (Figure 13), and strokes (Figure 14).
In states with high rates of diabetes,
teens were more likely to report intimate
partner violence. States with high rates of
stroke had more children exposed to adverse
childhood events. Potential causal pathways
are likely complex, including unobserved
factors and bidirectional relationships.
Access to health care was also poorer
in states with higher rates of obesity,
diabetes, and cardiovascular conditions.
Primary care shortages were more
common, and care was less affordable.
A W
ORD
ABO
UT M
ETHO
DS We examined how strongly
health outcomes correlated
with state statistics in five
domains that shape health:
health behaviors, the physical
and social environment, social
and economic factors, health
care, and public policies and
spending. The results, presented
in Figures 9 to 15, are based
on Spearman rank-order
correlation coefficients (rs),
which measure the degree to
which the state ranking for the
indicator (e.g., poverty) matches
the state ranking for the health
outcome (e.g., infant mortality).
Zero represents no association
between the two rankings, and
1.0 represents an exact match.
A positive correlation means
that a high rank on the indicator
is linked to a high rank on the
health outcome, or vice versa;
a negative correlation means
that a high rank on the indicator
is linked to a low rank on the
health outcome, or vice versa.
See Supplement 1: The Health of
the States: Spotlight on methods2
for more details on data sources
and methods and the rationale
for omitting certain results from
this report.
17
0
5
10
15
20
25
3.3%
11.2%
Stroke prevalence
3.1%
19.6%
Cerebrovascular mortality
4.2%
19.9%
Diabetes mortality
5.7%
9.6%
Diabetes prevalence
3%
16.7%
Adult overweight and obesity
FIGURE 16BUILT ENVIRONMENT IN TOP 10 AND BOTTOM 10 STATES FOR WALKABLE NEIGHBORHOODS (%)
Top 10 states (lowest disease rates)
Bottom 10 states (highest disease rates)
0
4
2
6
8
10
12
0.1%
8.9%
Cerebrovascular mortality
0.9%
7%
Angina or CHD* prevalence
0.1%
9.9%
Diabetes mortality
0.8%
7.4%
Adult overweight and obesity
FIGURE 16.2BUILT ENVIRONMENT IN TOP 10 AND BOTTOM 10 STATES FOR COMMUTING BY PUBLIC TRANSIT (%)
Top 10 states (lowest disease rates)
Bottom 10 states (highest disease rates)
0
20
10
30
40
50
60
19.8%
39.1%
Cerebrovascular mortality
*CHD = coronary heart disease
21.1%
50.1%
Angina or CHD* prevalence
22.7%
43.2%
Diabetes prevalence
20.3%
49.4%
Adult overweight and obesity
FIGURE 16.3BUILT ENVIRONMENT IN TOP 10 AND BOTTOM 10 STATES FOR PROXIMITY TO PARKS (% WITHIN HALF MILE)
Top 10 states (lowest disease rates)
Bottom 10 states (highest disease rates)
18
was 17.0 percent, compared with 10.8
percent in Top 10 states.
The convergence of household poverty
and area poverty was a particularly strong
predictor of these diseases. For example,
in Bottom 10 states with the highest
prevalence of strokes, 32.3 percent of the
and mortality rates from these diseases
were high, there were also very high
rates of household poverty, concentrated
neighborhood poverty, single-parent
households, and poor housing conditions.
For example, in Bottom 10 states with the
highest rates of strokes, the poverty rate
FIGURE 17AVOIDABLE HOSPITALIZATIONS IN TOP 10 AND BOTTOM 10 STATES (DISCHARGES PER 1,000) Bottom 10 states (highest disease rates)
Top 10 states (lowest disease rates)
0
20
40
60
80
50.3%
83.3%
Cardiovascular mortality
48.7%
81.2%
Heart attack prevalence
51.5%
80.2%
Diabetes prevalence
50.2%
82.0%
Adult overweight/obesity prevalence
$0
1K
2K
3K
$943
$2,255
Stroke prevalence
$920
$2,120
Cerebrovascular mortality
$1,083
$2,162
Heart attack prevalence
$1,148
$2,253
Diabetes mortality
$1,018
$2,206
Adult overweight and obesity
FIGURE 18INCOME SUPPORT PER CAPITA FOR PERSONS IN/NEAR POVERTY (INCOME LESS THAN 200% FPL) IN TOP 10 AND BOTTOM 10 STATES
Top 10 states (lowest disease rates)
Bottom 10 states (highest disease rates)
19
tax rates for tobacco and broader Medicaid
eligibilitye (Figure 15). The tobacco tax
was $2.68 per pack in Top 10 states with
the lowest cerebrovascular mortality
rates, more than four times the tax rate
in Bottom 10 states ($0.63 per pack);
Medicaid eligibility was 142.3 percent
and 65.9 percent of the Federal poverty
level, respectively. As seen in Figures
9–15, disease rates were often highest
in states that spent less (per poor person)
on services such as unemployment
benefits, income support, public assistance,
and social service workers. The Top 10
states spent twice that of Bottom 10 states
per capita on income support relative
to the size of the population in/near
poverty (earning less than 200 percent
of the poverty level) (Figure 18).
e. Refers to states’ Med
icaid income eligibility
limits for adultsparents
of dependent children
and other nondisabled
adults as a percent of
the Federal poverty
level, as of August 2014.
population lived in areas of concentrated
poverty,d more than twice that of Top 10
states (12.6 percent); the corresponding
percentages for poor residents living in
concentrated poverty were 55.0 percent
and 33.6 percent, respectively. Food
insecurity was more common in states
with higher rates of diabetes prevalence
and mortality (Figures 10–11), strokes
(Figure 14), and cerebrovascular mortality
(Figure 15). Adult incarceration was
also higher in states with higher rates of
strokes and cerebrovascular mortality.
States’ policies also correlated with
these conditions. For example, states
with lower rates of overweight and
obesity spent more per capita on mass
transit (Figure 9), and states with lower
cerebrovascular mortality had higher
d. Concentrated poverty is
defined as 20 percent or
more of the area popula
tion living with incomes
below the Federal
poverty level.
What The Data Affirm: The Takeaway
Obesity, diabetes, and cardiovascular disease — which together account
for enormous health burden and cost in the United States — are influenced
heavily by individual behaviors, such as lack of exercise and smoking.
But these conditions occur more commonly in places where residents live
in an unhealthy and unsafe environment, struggle with socioeconomic
challenges, and lack health insurance and access to affordable primary care.
The bottom line? The epidemics of obesity, diabetes, and heart disease
are not the result of unhealthy lifestyles alone; they are less common in
states and communities that offer better economic conditions for families,
better educational outcomes for children, widespread access to health care,
and communities and neighborhoods designed to encourage — and remove
barriers to — healthy living.
20
.
1. Woolf SH, Aron L, Chapman DA, et al. The
Health of the States: How U.S. States Compare
in Health Status and the Factors that Shape
Health—Summary Report. Richmond, VA: Center
on Society and Health, Virginia Commonwealth
University, 2016.
2. Woolf SH, Aron L, Chapman DA, et al. The Health
of the States: How U.S. States Compare in Health
Status and the Factors that Shape Health—
Spotlight on Methods. Richmond, VA: Center
on Society and Health, Virginia Commonwealth
University, 2016.
3. Krueger PM, Tran MK, Hummer RA, Chang VW.
Mortality attributable to low levels of education in
the United States. PLoS One. 2015;10:e0131809.
4. Brook RD, Rajagopalan S, Pope CA III., et al.
Particulate matter air pollution and cardiovascular
disease: an update to the scientific statement
from the American Heart Association. Circulation
2010;121:23312378.
5. Kaufman JD, Adar SD, Allen RW, et al.
Prospective study of particulate air pollution
exposures, subclinical atherosclerosis, and clinical
cardiovascular disease: The MultiEthnic Study of
Atherosclerosis and Air Pollution (MESA Air). Am J
Epidemiol. 2012;176:82537.
References
21
Will Monson, Rolf Pendall, Bryce Peterson,
Kathryn Pettit, Molly Scott, and Janine Zweig.
We also thank Stephanie Zaza, Centers
for Disease Control and Prevention, for
assistance in accessing data from the Youth
Risk Behavior Surveillance System (YRBSS)
and Robert Johnson, Vanderbilt University,
for biostatistical consulting. Other col-
leagues who gave us advice included Oscar
Arevalo, Nicklaus Children’s Hospital;
Elizabeth Bradley, Yale University; Ichiro
Kawachi, Harvard School of Public Health;
Matthew Penn, Public Health Law Program,
Centers for Disease Control and Prevention;
Robert Phillips, Jr., American Board of
Family Medicine; Christopher B. Swanson,
Editorial Projects in Education; Daniel
Taber, University of Texas Health Science
Center at Houston, School of Public Health;
Alan Ellis, Joseph Morrissey, and Kathleen
Thomas, University of North Carolina
Cecil G. Sheps Center for Health Services
Research; and Angela Kimball, National
Alliance on Mental Illness.
FUNDING
This project was funded by grant
number 71508 from the Robert Wood
Johnson Foundation.
Although any errors or omissions are
those of the authors only, we would like to
thank our Expert Advisory Panel, which
included Nancy Adler, Paula Braveman,
Debbie Chang, Ana Diez Roux, Neal
Halfon, David Kindig, Anna Schenck, and
Jonathan Showstack. We also appreciate
the advice we received from the staff of
the Robert Wood Johnson Foundation,
notably Matthew Trujillo, who served as
our program officer, and his predecessor,
Herminia Palacio.
We thank our colleagues at Virginia
Commonwealth University for their roles
in this study, including Sarah Blackburn
and Cassandra Ellison for graphic design,
layout, and dissemination of this report,
Lauren Waaland-Kreutzer for data verifi-
cation, and Jill Hellman, for administrative
support. We also thank Allison Phillips for
managing the first phases of this project
and Steven Cohen for providing advice on
demographic research methods.
We thank our colleagues at the Urban
Institute, especially Julia Isaacs for guiding
our analysis of spending data, but also
William Adams, Nan Astone, Richard
Auxier, Maeve Gearing, Linda Giannarelli,
Chris Hayes, Olivia Healy, Carl Hedman,
Carrie Heller, Ryan King, Carlos Martin,
Acknowledgments
This report is one of a series produced, in partnership with the Urban Institute, as part
of the Health of the States project — an initiative funded by the Robert Wood
Johnson Foundation (grant number 71508). For more information on the project, and
to view other reports in the series, visit societyhealth.vcu.edu.
Virginia Commonwealth University
VCU Center on Society and Health
830 East Main Street, Suite 5035
P.O. Box 980212
Richmond, Virginia 23298-0212
(804) 6282462
© Virginia Commonwealth University
Center on Society and Health, 2017
Top Related