Nutrition Environment Measures Survey (NEMS)

1

Nutrition Environment Measures Survey
(NEMS) Overview

2

• Why study nutrition environments?
• Obesity is epidemic
• Psychological social factors dont fully
  explain diet physical activity
• Food environments are not well understood
• Need to understand the food environment in order
  to develop interventions

3
Environmental Causes of Obesity
Environments are believed to be important among
the multiple levels of determinants of nutrition
and physical activity

• Increased eating
• Decreased energy expenditure

4
The Rationale for NEMS Why do a MEASUREMENT
study? ? In order to do good research, we need
reliable and valid measures ? Good measures of
nutrition environments dont exist ? We need to
understand the practical side of measures (time,
efficiency)
5
Model of Community Nutrition EnvironmentsGlanz,
Sallis, Saelens Frank 2005
This model blends public health, health
psychology, urban planning perspectives ?
Guides developments of measures research
priorities
6

• 4 types of nutrition environments
• Community
• Consumer
• Organizational
• Informational
• Moderating mediating pathways hypothesized
• Less studied may have broad effects

7

• Community Consumer Nutrition Environments
• Community nutrition environments
• Type location of food outlets
• Accessibility (e.g., hours, drive-thru)
• Consumer nutrition environments
• Availability of healthful food choices
• Pricing, promotion, placement
• Information availability

8

• This model is a starting point
• Complex research practice
• area
• Greater priority is needed for
• nutrition environments

9
Nutrition Environment Measures Study (NEMS)

Funded by the Robert Wood Johnson Foundation
10

• Aims of NEMS
• Develop measures of nutrition environments
  retail food service outlets (stores,
  restaurants)
• Test the inter-rater reliability and test-retest
  reliability of nutrition environment measures
• Examine sampling and generalizability issues

11
The Most Important Measurement Concepts Validity
Reliability
12
Validity Whether an instrument measures what it
proposes to measure Measures reflect true
differences in the things they intend to measure
13

• Types of Reliability Examined in NEMS
• Inter-rater reliability (equivalence)
• Test-retest reliability (stability)

14

• Reliability in NEMS 1
• Inter-rater reliability
• 2 raters go to same store/restaurant, same day
  (same time) ? ? Do they get the same results?

15
Reliability in NEMS 2 2. Test-retest
reliability The same rater goes to same
store/restaurant, one week apart ? Does he/she
find the same results? This assumes stores dont
change that fast, but were not sureIt depends
on measures that arent too subject to random
error
16
Test-Retest Reliability and Inter-Rater
Reliability
Rater 1
Rater 2
TIME 1
Inter-Rater Reliability
Test-Retest Reliability
Test-Retest Reliability
Rater 1
Rater 2
TIME 2 (T11 week)
17

• Phases of the study
• Pre-test
• Preliminary work
• develop measures, try them out, improve them
  for formal research purposes
• Where? Atlanta neighborhoods -- Decatur (hi
  walk) Toco Hills (lo walk) neighborhoods

18

• Phases of the study (contd)
• Main measurement study
• Collect data to allow us to calculate
  test-retest reliability inter-rater reliability
  in 4 neighborhoods around schools
• Where? High/low walk High/low SES
• neighborhoods in metro Atlanta
• Hi walk, hi SES
• Hi walk, low SES
• Lo walk, hi SES
• Lo walk, lo SES

19

• Selection of Communities Identification of Food
  Outlets
• Communities selected using maps, census data, GIS
• Outlets enumerated by project staff online
  directories, business directories, health
  department, etc.

20

• Measures of Nutrition Environments in Stores
• lt Grocery Stores Convenience Stores gt
• Availability (of healthful choices)
• Prices (compare healthy to less healthy grocery
  to convenience stores)
• Quality (for fresh produce)

21
Measures of Nutrition Environments in Stores lt
Grocery Stores Convenience Stores gt Core
Categories of Foods Milk Ground Beef Baked
Goods Fruits Hot Dogs Bread Vegetables Froze
n Dinners Baked Chips Fruit Juice
22
Grocery Stores Consumer Nutrition Environment
Measures
23
(No Transcript)
24
Convenience Stores Consumer Nutrition Measures
25
Measures of Nutrition Environments in
Restaurants lt Fast-Food Sit-Down Restaurants gt

• Sources of Information
• Internet
• Menu
• Visit, observation
• Interview manager

26
Measures of Nutrition Environments in
Restaurants lt Fast-Food Sit-Down Restaurants gt

• Availability (of healthful choices)
• Prices (compare healthy vs. less healthy
  fast-food vs. sit-down)
• Promotion, Information
• Facilitators Barriers
• Kids Menus

27
Restaurant Measures
28
NEMS Raters in the Field
29
Cover Page
30
Fruit
31
Vegetables
32
Hot Dogs
33
Baked Chips
34

• Findings

35
88 Stores (90.6 completion rate)

• 24 grocery
• 64 convenience
• 16-27 stores per neighborhood

• 301 Restaurants
• 129 SDRs in High-Walk, High-Income Area
• Sample of 40
• 217 restaurants assessed
• 102 fast food (99 completion rate)
• 115 sit-down (100 completion rate)

36
High Income/ High-Walkability
37
High Income/Low-Walkability
38
Inter-Rater Reliability of NEMS Store Observations
Variable/Indicator Inter-Rater Reliability (2 raters, same day) agreement Kappa/V a
Any fruit availability 96.47 .93
Any vegetables availability 100 1.00
Baked chips 96.47 .92
Lean ground beef 98.82 .96
100 Whole grain bread 92.94 .83
Skim/low-fat milk 100 1.00
Hot dogs (regular vs. fat-free) 100 1.00
Reduced calorie frozen dinner 100 1.00
Low-fat baked goods 95.29 .88
a Cramers V statistic used when Kappa could not
be computed due to asymmetric rater response
dimensions
39
Test-Retest Reliability of NEMS Store Observations
Variable/Indicator Test-Retest Reliability (1 rater, 2 weeks apart) agreement Kappa/V a
Any Fruit availability 92.68 .85
Any Vegetables availability 96.34 .91
Baked chips 95.12 .89
Lean ground beef 98.78 .96
100 Whole grain bread 90.24 .75
Skim/low-fat milk 97.56 .95
Hot dogs (regular vs. fat-free) 98.78 .95
Reduced calorie frozen dinners 98.78 .96
Low-fat baked goods 93.90 .84
a Cramers V statistic used when Kappa could not
be computed due to asymmetric rater response
dimensions
40
Grocery Stores vs. Convenience Stores
Availability of Fruit Vegetables
Plt .001
Plt .001
41
Grocery Stores vs. Convenience Stores
Availability of Fat-Free Hot Dogs Baked Chips
Plt .01
Plt .001
42
High- vs. Low-Income Neighborhoods Availability
of Fruits Vegetables
Plt .01
Plt .01
43
High- vs. Low-Income Neighborhoods Availability
Fat-Free Hot Dogs Baked Chips
Plt .01
n.s., trend
44
Shelf Space Skim Milk vs. Full-Fat Milk 40
skim milk Higher in GS High SES areas
45
Cost Comparisons

• Fruits
• Bananas
• .47/.62 76 GS/CS (plt.001)
• Milk
• .99 ratio skim/full fat (NS)

46
Cost Comparisons

• Hot Dogs
• 124 of regular for lean franks
• Ground Beef
• 147 of regular for lean meat
• Chips
• 131 of regular for low-fat
• Juice
• 153 of juice drink for 100 juice

47
NEMS Composite Mean Scores for Healthy Nutrition
Environments in Stores
Grocery Stores (n24) Convenience Stores (n61)
Availability 17.33 3.54
Price 0.13 1.54
Quality 5.13 .077
Total 22.58 5.85
Maximum possible total score is 50 points
48
NEMS Composite Mean Scores for Healthy Nutrition
Environments in Stores
Price
Total
Maximum possible total score is 50 points
49
NEMS Composite Mean Scores for Stores By
Neighborhood SES
High Income (n44) Low Income (n41)
Availability 10.23 4.44
Price 0.30 2.05
Quality 2.61 1.34
Total 13.14 7.83
Maximum possible score is 50 points
50
NEMS Composite Mean Scores for Stores by
Neighborhood SES
Price
Total
Maximum possible total score is 50 points
51
Time for Completing Measures

• Convenience stores
• mean 14.4 min range10-18 min
• Grocery Stores
• mean 41.8 min range 30-66 min
• Restaurant site visits average 11.5 minutes
  (9-35)
• Menu reviews average 35 minutes

52
Limitations Other venues where food is sold
not included May have left out some
important variables
53
What we do do not know

• Environmental vs. individual/social determinants?
• Distribution of unhealthy environments (SES,etc.)
• How much environmental change is needed?

54

• Lets build a better mousetrap

55
Acknowledgments Jim Sallis Larry Frank Brian
Saelens Terry Conway Esther Friedman Linda
Schuessler Kristi Maxwell Michelle
Carvalho Monique Young Molly Dowling Margaret
Clawson Jim Chapman Melanie Baptiste Funding
Support