Assessing and Addressing Inequities in Community Nutrition in Washington State Marilyn Sitaker, WA DOH Public Health Nutrition 1/13/2011.

Assessing and Addressing Inequities in Community

Nutrition in Washington State

Marilyn Sitaker, WA DOH Public Health Nutrition 1/13/2011

What is Health Equity?

Health Equity is the absence of differences in health between groups with greater and lesser levels of social advantage

Health equity is necessary for individuals & groups to participate in, and benefit from, social and economic development.

Health equity is a conscious process requiring effort

Today’s Lecture Topics

1. How socioeconomic conditions are linked to inequalities in health status & health outcomes

2. How to measure constructs in the health equity model at the state level

3. How researchers link inequities in access to healthy foods to differences in nutrition behaviors among social groups

4. Intervention strategies

5. Department of Health initiatives (time permitting)

1. How socioeconomic conditions are linked to inequalities in

health status & health outcomes

Key ideas from “Bad Sugar”

“Reaching for a Healthier Life”Facts on Socioeconomic Status & Health in the US

(1) Socioeconomic status has a big impact on everyone’s health. Premature death is 3 times more likely for those who live in poverty compared to those who are most privileged.

(2) Throughout our lives, access to socioeconomic resources affects our chances for living a healthy life. The conditions we live in during childhood affect our health throughout our lives.

(3) Health care is important, but accounts for only a small portion of health disparities. Social determinants are more important in determining whether we fall ill in the first place.

http://www.macses.ucsf.edu/downloads/Reaching_for_a_Healthier_Life.pdf

(4) Each step up the social ladder provides greater access to social and physical environments that make it easier to engage in healthy behaviors, (e.g., safe places to walk and access to healthier foods). Each step down, greater exposure to potential risks (pollution & unsafe neighborhoods).

(5) Work conditions contribute to health & health disparities. Low-wage jobs may involve shift work and physical hazards, low control over how and when tasks are done, job insecurity, and conflicts between family obligations and work requirements.

(6) Exposure to extreme and prolonged “toxic” stress is more common lower on the social ladder. Persistent stressors--financial insecurity, interpersonal disputes, work-induced exhaustion, chronic conflict-- are recorded in the body.

“Reaching for a Healthier Life”Facts on Socioeconomic Status & Health in the US

http://www.macses.ucsf.edu/downloads/Reaching_for_a_Healthier_Life.pdf

Conceptual Model created by the World Health Organization Commission on Social Determinants of Health http://www.who.int/social_determinants/resources/csdh_framework_action_05_07.pdf

How social conditions influence health equity

2. How to measure the link between disparities in access to

social resources and health outcomes

Summary Measures to Compare Disparities

By Education & IncomeAbsolute measures compare the difference in risk between the highest and lowest group:

11 - 5 = 6%

Relative measures use a ratio or risk in the highest & lowest income groups:

11 ÷ 5 = 2.2

11

7

5

0 5 10 15 20

<$35,000

$35,000-$74,999

$75,000 or more

Age-Adjusted Percent

Ho

use

ho

ld In

com

e

Diabetes Among Adults by Income in Washington, 2007-2009

Source: Washington Behavioral Risk Factor Survey

Disparities in Risk Factors & Chronic Diseases among Washington Adults by

IncomeChronic Disease Risk Factors Absolute RatioSmoking 2007-2009 1.7% 3.1Insuffi cient physical activity 2007 & 2009 7.0% 1.2Insuffi cient fruit & vegetables 2007 & 2009 8.4% 1.1Chronic Disease Risk ConditionsObesity 2007-2009 9.5% 1.4Hypertension 2007 & 2009 9.8% 1.5High Cholesterol 2007 & 2009 6.1% 1.2Chronic Disease PrevalenceAsthma 2007-2009 4.1% 1.5CVD 2007-2009 5.5% 2.4Diabetes 2007-2009 6.0% 2.3

Adults with household incomes above $75K compared to those <$35K

Data Source: WA Behavioral Risk Factor Surveillance SystemNote: All differences between highest and lowest income group are statistically significant.

How many people are affected?

EducationPrevalence of Obesity

State Population

age 25+

Number Obese (population

X percent)

Number affected if same prevalence as college

graduatesExcess Cases

HS or Less 31.7% 1,480,000 470,000 300,000 170,000

Some College 31.9% 1,530,000 470,000 310,000 160,000

College Graduate 21.1% 1,330,000 270,000 270,000 0

Total 4,350,000 1,210,000 880,000 330,000

Obesity Diabetes Smoking Hypertension

Total Number of Excess Cases 330,000 110,000 460,000 260,000

High School; 34.0%

Some College; 35.2%

College Grad; 30.5%

Source: Washington BRFSS 2006-2008

Physical Activity by SEP; Access to Local Outdoor Recreation by Socioeconomic

PositionThe less education a person has, the less likely it is that he or she lives near a public park, playground, trail or school recreational facility.

Less educated adults are also less likely to use nearby recreational facilities, & less likely to get enough physical activity.

Influence of neighborhood features on physical activity, all adults in

Washington 2005

0 20 40 60 80 100

Gets enough physicalactivity

Uses park, playground,school or trail

Lives near park,playground, school or trail

Percent

College or more

Some college

High school orless

Directory of Social Determinants of Health at the Local Level

University of Michigan SPH project funded by the CDC. Developers had expertise in diverse areas.

Directory lists current data sets that can be used to address SDOH. Data sets organized in 12 dimensions of the social environment.

Each dimension is subdivided into various components.

Source; Hillemeier M.M., J. Lynch, S. Harper, and M. Casper. 2003. "Measuring contextual characteristics for community health." Health Services Research 38(6 part 2):1645–717.

12 Dimensions of social context

Economic Dimension

This table presents the components and indicators of the economic dimension. Nine economic components are identified:

1. Income 2. Wealth 3. Poverty 4. Economic Development 5. Financial Services 6. Cost of Living 7. Redistribution 8. Fiscal Capacity 9. Exploitation

Source; Hillemeier M.M., J. Lynch, S. Harper, and M. Casper. 2003. "Measuring contextual characteristics for community health." Health Services Research 38(6 part 2):1645–717.

Components and Indicators Data Sources and Notes

1. Income: Earned income

Median and per capita annual income Census Bureau

Mean hourly and annual wage Bureau of Labor Statistics Data by occupation available in downloadable Excel files.

Hourly wage, union, and nonunion workers

Union Membership and Earnings Data Book (http://www.bna.com/bnaplus/labor/ laborrpts.html). Separate tables for public and private sector workers and for manufacturing and nonmanufacturing workers. Customized reports available for any or all years since 1983.

Per capita personal income Bureau of Economic Analysis Downloadable compressed comma–separated–value files.

Income: Disposable income

Median and per capita Effective Buying Index

Demographics U.S.A. (http://www.tradedimensions.com/ p_demographics.html). Effective Buying Index represents money income minus taxes. Data available on CD–ROM.

Income: Income distribution

Gini coefficient of income inequality; 90%ile/10%ile ratio

Census Bureau

Indicators & Measures: Income

Harvard Geocoding Project: Measuring Socioeconomic Position (SEP)

Key domains: Occupational class: affects health via occupational

hazards and income/standard of living

Educational attainment: reflects childhood SEP and future economic prospects, also knowledge & health literacy

Income & subsidies: affects standard of living

Wealth: referring to accumulated assets

Relative social ranking: “status” and “prestige”

Source: Public Health Disparities Geocoding Project

Assessments can be made within socioeconomic class domains at the individual, household, and area or neighborhood level.

Socioeconomic data can be measured at key points in the lifecourse -- in utero, infancy, childhood, and early, middle, and late adulthood.

Composite measures can be constructed to combine information. For example, the Townsend index consists of % unemployment, % renters, % not owning a car, and % crowding.

Area Based Measures of SEP

This economically depressed area in Boston's Chinatown, turned out to be characterized as a highly working class, poor, low income area with high unemployment and few expensive homes.

This one house in Beacon Hill looked like it was -- and turned out to be -- in a fairly affluent area: over 75% professionals, low poverty, high income, low unemployment, and lots of expensive homes.

Comparing two Boston neighborhoods

3. Evidence for impact of inequity in the distribution of social

resources that support healthy eating

Assembling a Mosaic of Evidence

“The community nutrition environment may explain some of

the racial, ethnic and socioeconomic disparities in nutrition and

health such as the increasing prevalence of overweight in low

income children. Supermarkets...are less common in lower

income and minority neighborhoods than in other

neighborhoods…recent evidence links access to supermarkets

with…fruit and vegetable intake among African American

adults…”

The role of the built environments in physical activity, eating and obesity in childhood, Sallis J, Glanz, K. www.futureofchildren.org, vol 16 (1), 2006.

“Supermarkets...are less common in lower income and minority neighborhoods”

A study of access to food markets and restaurants by neighborhood wealth (median HH income) in MS, NC, MD and MN showed that wealthy neighborhoods had 3 times as many grocery stores as poor neighborhoods. Supermarkets were 4 times more common in white neighborhoods compared to black neighborhoods (Moorland et al, Am J Prev Med 2002; 22(1)

Spatial regression analysis of average distance to the nearest supermarket in 869 Detroit neighborhoods showed that distance to nearest supermarket was about the same in wealthier neighborhoods, regardless of racial makeup. Among poor neighborhoods, those with high proportion of African Americans were 1.1 miles further from the nearest market than white neighborhoods. (Zenk et. al, Am J Pub Hlth 2005 95(4)

“…access to supermarkets linked to…fruit and vegetable consumption…”

Analysis of 10,623 food frequency questionnaires with geocoded home

address compared with geocoded location of local supermarkets showed

that for each additional supermarket in the neighborhood, fruit and

vegetable intake increased by 31% for blacks and 11% for whites.

Morland, et. al, Am J Pub Hlth 2002; 92(11)

A study of fruit and vegetable consumption among food stamp participants

showed that households living more than 5 miles from their principal store

consumed less fruit than those living within a mile of their store

Rose, et. al, Pub Hlth Nutrition 2004, 7 (8)

4. Disparities in nutrition behaviors and environments

that support healthy eating in Washington State

Washington: Disparities in Eating F&V

Adults with the lowest incomes & educational level are less likely to eat enough fruit and vegetables.

Certain racial groups are also less likely to meet dietary guidelines.

2126

30

1924

3226

28

3228

2421

0 20 40

Less than $35,000

$35,000 to $74,999

$75,000 or more

High school or less

Some College

College graduate or more

White*

Black*

Asian*

Pacific Islander*

American Indian/Alaska Native*

Hispanic

Age-Adjusted Percent

Eats F&V 5 times Daily, by race, income, and education Washington State, 2007-2009

* Non-HispanicSource: WA Behavioral Risk Factor Survey (BRFSS), 2007-2009

Race / EthnicityEducation

Household Incom

e

Likelihood of being food insecure, taking multiple causal factors into account

Causal Factors:

AgeEducation Income

Race/ethnicityMarital StatusSex

Smoking StatusHealth Status

1.5

2.22.3 2.6

1.8 1.6

0.9

1.5 1.4

0

1

2

3

4

Od

ds

Rat

io

Regression Analysis Results: Relative Odds Ratio for Food Insecurity

Note: Interactions occur between Poor Health & Low Income; Poor Health & Current Smoking and Hispanic

Income & Age are the Strongest Determinants of Food Insecurity

Regression Analysis Results: Relative Odds of Experiencing Food Insecurity: Annual Household Income and Age

38

18

25

11

1515

37

0

10

20

30

40

50

60

70

Less than$25,000

$25,000-34,999 $35,000-49,999 $50,000-74,999 Age 20-34 Age 35-44 Age 45-54 Age 55-64 Age 65-74

Odds

Ratio

• Income <$25,000/year: 38 times more likely than income $75,000+

• Ages 20-44: 15 times more likely to be food insecure than ages 75+.

Washington: Trends in Disparities in Eating F&V

0

10

20

30

40

50

60

70

80

90

1990 1995 2000 2005 2010

Ag

e A

dju

sted

Per

cen

t

Year

Eats fruits and vegetables < 5 times a day among Washington Adults, by education 1994-2009*

HS or less

Some collegeCollege grad

Source: Washington Behavioral Risk Factor Surveillance System; date represents the midpoint of 2 year averages of data collected in alternate years.

0

2

4

6

8

10

12

14

1990 1995 2000 2005 2010

Ag

e A

dju

sted

Per

cen

t

Year

Excess risk of poor diet by education, comparing HS education or less to college

graduate

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

1990 1995 2000 2005 2010R

isk

Rat

ioYear

Relative risk of poor diet by education, comparing HS education or less to college

graduate

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

1990 1995 2000 2005 2010

Ris

k R

ati

oYear

Relative risk of obesity by education, comparing HS education or less to college graduate

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

1990 1995 2000 2005 2010

Ag

e A

dju

ste

d P

erc

en

t

Year

Excess risk of obesity by education, comparing HS education or less to college graduate

0

5

10

15

20

25

30

35

1990 1995 2000 2005 2010

Ag

e A

dju

ste

d P

erc

en

t

Year

Obesity among Washington Adults, by Education, 1990-2008

HS or less Some college College grad

9.425.6

9.425.6

Source: Washington Behavioral Risk Factor SurveyNote: Only among adults who answered survey in English.

Slope= 1.16

Slope= 0.5

Washington: Trends in Disparities in Obesity

Deep green = Washington Counties

most likely to have insufficient F&V

consumption

Insufficient F&V consumption BRFSS, 2005-2007

4. Intervention strategies promoted in Reaching for a

Healthier Life

Policies to Promote Health Equity1. Policies that Affect the Ladder 2. Policies that Blunt Adverse Consequences

5. Initiatives within the Department of Health Initiatives (Community

Wellness and Protection)

Partners in Action Websitewww.wapartnersinaction.org

Healthy food/drink availability

Limit unhealthy food/drink availability

Farm to institution, including schools, worksites, hospitals, and other community institutions

Menu labeling

Support breastfeeding through policy change and maternity care practices

Promote Physical Activity

Safe, attractive accessible places for activity

City planning, zoning and transportation

Require daily quality PE in schools

Require daily physical activity in afterschool/childcare settings

Safe routes to school

0 2 4 6 8 10 12 14

4

1

2

1

1

1

2

7

1

1

1

3

8

12

2

2

3

State & Local Policies Adopted in Washington State*Based on MAPPS Strategies, 2005-2010

State Policies = 12 Local Policies = 42

*These are policies in which DOH was directly involved; it does not reflect all Washington policies enacted during this period.

Paula Braveman: Thoughts on Health Inequities

Systematic differences in health or health determinants that are plausibly influenced by social policy are health inequities if they

a) Occur between groups with different social position (place in the hierarchy according to power, wealth, prestige)

b) Place groups already at social disadvantage at even greater disadvantage due to poor health

You do not need to attribute causation or prove that the disparity is avoidable if social policies were changed, as long as the impact is plausible.

Braveman, 2004, Health Policy and Development 2(3) 180-185

Thank You!

Marilyn Sitaker, MPHChronic Disease Prevention Unit

Lead Epidemiologist and Evaluation Coordinator

(360) [email protected]