Fruits and Vegetables in South Phoenix

1
Fruits and Vegetables in South Phoenix

• Kai-Ning Khor, MPH
• June 12, 2008

2
Overview

• Learning Objectives
• Background
• Survey Methods
• Data Analysis Methods
• Results
• Conclusion
• Discussion
• Challenges in survey methods
• Challenges in data analysis methods
• Future

3
Learning Objectives

• Participants will be able to
• Describe the project Arizonas Healthy Weight
  Action Learning Collaborative completed.
• Recognize the lessons Arizonas team learned
  regarding the NEMS.
• Identify ways the fruit and vegetable
  availability will be affected with the new WIC
  food package

4
Background

• Nutrition Environment Measures Survey (NEMS)
• Observational measures of community nutrition
  environments in stores
• Measures focus on availability of healthful
  choices, prices and quality
• Goal To determine whether the availability and
  quality of healthy foods in South Phoenix affects
  the BMIs of WIC participants in that area

5
Background South Phoenix

• Area 80 square miles
• From 40th Street to 35th Avenue, and from Buckeye
  to Baseline
• Population has
• Limited resources, low socioeconomic status
• Low levels of educational attainment
• Health disparities
• High levels of the feto-infant mortality
• Source Pregnancy Risk Assessment Monitoring
  System (PRAMS) survey and Perinatal Periods of
  Risks (PPOR) analyses, 2004

6
Survey Methods

• Purpose of survey was to assess whether healthy
  options of various foods were available, their
  quality, and their costs

• Foods included
• Milk
• Fruit
• Vegetables
• Ground beef
• Hot dogs
• Frozen dinners

• Baked goods
• Beverages
• Bread
• Baked chips
• Cereal
• Tortillas

7
Survey Methods, continued

• Two students/volunteers surveyed each store 1-2
  times separately
• 5 groups, 2 surveyors/group
• Permission obtained from managers
• Survey period August-November 2007
• 76 stores surveyed, 238 surveys total
• 66 convenient stores, 10 grocery stores
• 17 WIC stores, 59 non-WIC stores

8
Data Analysis Methods

• Two independent sample T tests
• Levenes Tests for Equality of Variances
• Geographic Information Systems (ArcGIS v 9.2)

9
Results Pre-Pregnancy BMI of WIC women in South
Phoenix

• The range for average BMI by store was 25.7 to
  26.5.  
• Fairly homogeneous population

10
(No Transcript)
11
Results Grocery vs. Convenience Stores

• On average, grocery stores offered a wider
  variety of fruits than convenient stores

12
Results Grocery vs. Convenience Stores

• On average, grocery stores offered a wider
  variety of vegetables than convenient stores

13
Results Grocery vs. Convenience Stores

• On average, grocery stores offered better
  quality fruits and vegetables than convenient
  stores

14
Results WIC vs. Non-WIC Stores

• On average, WIC stores offered a wider variety
  of fruits than non-WIC stores

15
Results WIC vs. Non-WIC Stores

• On average, WIC stores offered a wider variety
  of vegetables than non-WIC stores

16
Results WIC vs. Non-WIC Stores

• On average, WIC stores offered better quality
  fruits and vegetables than non-WIC stores

17
Conclusions

• BMIs of WIC participants in surveyed South
  Phoenix area fairly homogeneous
• Fruit and vegetable availability and quality
  differed throughout surveyed area
• Availability and quality of fruits and vegetables
  in South Phoenix did not affect the BMIs of WIC
  participants in that area
• Possible confounding factors consumption, foods
  actually purchased, transportation, other sources
  of fruits and vegetables in area (farms)

18
Conclusions, continued

• New WIC food package would help to increase the
  varieties of fruits and vegetables available,
  their quality, and their affordability to WIC
  participants

19
Challenges in survey methods

• Information different between grocery and
  convenience stores
• Possibly use different surveys for each
• Consistent training for surveyors

20
Challenges in analysis methods

• Analyzing qualitative data
• Performing inter-rater reliability and
  test-retest reliability tests for validity
• Not feasible because of
• Inconsistencies in data collection
• Number of surveys per store varied
• Inconsistencies in data entry

21
Future

• More training by Emory University
• Overlay population density map to assess where
  stores are and where residents live
• Enhance and expand survey
• Survey area with similar demographic population
  to make comparisons
• Include more grocery stores and convenience
  stores
• Include restaurants in area

22
Future, continued

• Survey residents for consumption and purchase
  behaviors before and after new WIC food package
  is implemented
• Evaluate whether the new WIC food package would
  help to increase sales
• Assess whether sales differ between WIC and
  non-WIC stores

23
Questions/Comments?

• Kai-Ning Khor, MPH
• Chronic Disease Epidemiologist
• E-mail khork_at_azdhs.gov
• Phone [phone]
• Fax [phone]