Research article
Nutrition Environment Measures Survey in Stores (NEMS-S): Development and Evaluation

https://doi.org/10.1016/j.amepre.2006.12.019Get rights and content

Background

Eating, or nutrition, environments are believed to contribute to obesity and chronic diseases. There is a need for valid, reliable measures of nutrition environments. This article reports on the development and evaluation of measures of nutrition environments in retail food stores.

Methods

The Nutrition Environment Measures Study developed observational measures of the nutrition environment within retail food stores (NEMS-S) to assess availability of healthy options, price, and quality. After pretesting, measures were completed by independent raters to evaluate inter-rater reliability and across two occasions to assess test–retest reliability in grocery and convenience stores in four neighborhoods differing on income and community design in the Atlanta metropolitan area. Data were collected and analyzed in 2004 and 2005.

Results

Ten food categories (e.g., fruits) or indicator food items (e.g., ground beef) were evaluated in 85 stores. Inter-rater reliability and test–retest reliability of availability were high: inter-rater reliability kappas were 0.84 to 1.00, and test–retest reliabilities were .73 to 1.00. Inter-rater reliability for quality across fresh produce was moderate (kappas, 0.44 to 1.00). Healthier options were higher priced for hot dogs, lean ground beef, and baked chips. More healthful options were available in grocery than convenience stores and in stores in higher income neighborhoods.

Conclusions

The NEMS-S tool was found to have a high degree of inter-rater and test–retest reliability, and to reveal significant differences across store types and neighborhoods of high and low socioeconomic status. These observational measures of nutrition environments can be applied in multilevel studies of community nutrition, and can inform new approaches to conducting and evaluating nutrition interventions.

Introduction

Social and built environments are believed to exert important influence on individuals’ access to affordable, healthful food.1, 2, 3, 4, 5 Understanding of healthy nutrition environments, while growing, is limited by the lack of reliable and valid measures of these environments. To advance science, and to inform public health policy, there is a need for well-defined concepts and valid, reliable measures of nutrition environments. Both “community” and “consumer” nutrition environments can affect food choice. The community nutrition environment is comprised of the number, type, location and accessibility of food outlets such as grocery stores. The consumer nutrition environment is what consumers encounter in and around places where they buy food, such as the availability, cost, and quality of healthful food choices.6

The presence of food stores, and the availability of healthful products in those stores, appear to be important contributors to healthy eating patterns among neighborhood residents.7 Racial and ethnic disparities in access to supermarkets, that typically have good availability of healthful foods, have been documented.8, 9 Black Americans’ fruit and vegetable intake was considerably higher when they had more supermarkets in their census tracts,10 and proximity to a supermarket was favorably associated with the diet quality of pregnant women11 and with lower prevalence of obesity and overweight.12 Low-income women who shopped at supermarkets and specialty stores consumed more fruits and vegetables than those who shopped at independent grocers.13 Thus, the community nutrition environment appears to affect individual food choice and may affect long-term health.

Studies of the consumer nutrition environment within stores, particularly regarding availability, quality, and price, also reveal disparities and associations with dietary intake. Horowitz et al.14 found significantly lower availability of five foods recommended for diabetics in less-affluent and ethnic minority neighborhoods, and findings of lower availability of healthful foods in low-income and high-minority neighborhoods were replicated.15 The higher costs of more healthful foods is a concern.16 Jetter and Cassady17 found that most people in their study had access to healthful foods, but the healthier market basket was 35% to 40% more expensive than a standard market basket because of the higher costs of whole grains, lean ground beef, and skinless poultry. Cheadle et al.18 found community-level associations between grocery store environments and individual dietary practices, with the most significant relationships between low-fat milk consumption and the proportion of shelf space for skim, 1% fat, and 2% fat milk.

Despite the increased interest in nutrition environments, little progress has been made in devising reliable and valid measurement tools. Nearly 2 decades ago, Cheadle et al.19 assessed the proportion of shelf space devoted to healthful alternatives, such as reduced-fat milk, poultry and fish, and 100% whole wheat bread. They found high inter-rater reliability (0.73 to 0.78) and test–retest reliability ranging from 0.44 to 1.00. Horowitz et al.14 measured availability of five diabetic-recommended foods in grocery stores and reported excellent inter-rater reliability ranging from 0.94 to 1.00. Other published reports have been less clear about the rigor of their methods or did not report reliability of the measures.15, 17, 20

Psychometrically sound measures are needed to obtain accurate and reliable estimates of the relationship between nutrition environments and individuals’ dietary intake, as well as to evaluate change in nutrition environments secondary to intervention. The purpose of the present study was to develop observational measures of nutrition environments in retail stores, evaluate measure reliability, and examine differences in nutrition environments across different types of stores and between higher and lower income neighborhoods.

Section snippets

Methods

The Nutrition Environment Measures Study (NEMS) developed and evaluated nutrition environment measures for retail stores (NEMS-S, described here) and restaurants (NEMS-R, described in a separate paper).21

Description of Sample and Response Rates

A total of 88 stores (24 grocery, 64 convenience) were identified in the four study neighborhoods, ranging from 16 to 27 per neighborhood. A 100% completion rate for all three sets of measures was achieved in the grocery stores; however, three convenience stores declined to participate (for a 95.3% completion rate) and two other convenience stores refused second visits, yielding a net 90.6% rate for completing all measures at stores. This resulted in 24 grocery stores and 61 convenience stores

Discussion

The NEMS-S food store environment measures developed and evaluated in this study had high inter-rater and test–retest reliabilities and provide support for the construct validity of the measures. Because the indicator foods were carefully selected based on authoritative guidelines and recommendations, face validity of the measures is also affirmed. These measures provide an evaluation of food stores available in specific locations, so they assess aspects of the community nutrition environment.

References (40)

  • T. Pearson et al.

    Do “food deserts” influence fruit and vegetable consumption?—A cross-sectional study

    Appetite

    (2005)
  • S.L. Booth et al.

    Environmental and societal factors affect food choice and physical activity: rationale, influences, and leverage points

    Nutr Rev

    (2001)
  • S.A. French et al.

    Environmental influences on eating and physical activity

    Annu Rev Public Health

    (2001)
  • Health and behavior: the interplay of biological, behavioral, and societal influences

    (2001)
  • J.F. Sallis et al.

    Ecological models of health behavior

  • K. Glanz et al.

    Healthy nutrition environments: concepts and measures

    Am J Health Promot

    (2005)
  • K. Glanz et al.

    Strategies for increasing fruit and vegetable intake in grocery stores and communities: policy, pricing, and environmental change

    Prev Med

    (2004)
  • S.N. Zenk et al.

    Neighborhood racial composition, neighborhood poverty, and the spatial accessibility of supermarkets in metropolitan Detroit

    Am J Public Health

    (2005)
  • K. Morland et al.

    The contextual effect of the local food environment on residents’ diets: the Atherosclerosis Risk in Communities Study

    Am J Public Health

    (2002)
  • Cited by (555)

    • Food environment

      2023, Encyclopedia of Human Nutrition: Volume 1-4, Fourth Edition
    View all citing articles on Scopus
    View full text