Effects of Individual and Environmental Factors on Intervention Efficacy and Participant Retention in the American Cancer Society's Study of Nutrition and Physical Activity Open Access

Cross, Di (2009)

Permanent URL: https://etd.library.emory.edu/concern/etds/s4655h13d?locale=en


Observational studies have demonstrated positive associations between individual physical activity (PA) and environmental factors like increased access to parks, safer streets, and dense, mixed use of surrounding land. Similarly, greater fruit and vegetable consumption (FVC) has been observed to be associated with increased availability or greater variety of healthy foods. However, a question of temporality remains unanswered as studies documenting such associations have largely employed cross-sectional designs. The few longitudinal studies that have been performed may still leave that question unanswered, as the same associations would arise if individuals selected their environments based on desired behaviors rather than if the environment caused the behavior.

This dissertation aimed to address the question of temporality by examining environmental-level characteristics with respect to loss to follow-up and differential efficacy in a study of an intervention designed to modify PA and FVC behaviors. This was accomplished using baseline, and 3-, 6-, and 12-month follow-up data from the American Cancer Society's Nutrition and Physical Activity (NuPA) Study - a randomized controlled trial of a low-intensity, telephone-administered counseling intervention - and commercially-available databases of grocer and park locations, and street connectivity.

The greatest intervention effect on FVC was observed among participants with the least access to grocers. The greatest intervention effect on PA was observed among participants with the least access to parks. No consistent difference in intervention effect was found across levels of street connectivity. No characteristics of the nutrition or physical activity environment were associated with loss to follow-up.

Our findings suggest that participants with the least access to resources may benefit the most from this low-intensity counseling intervention. This reveals a more complex picture of the relationship between environment and behavior than previously thought and suggests that even if access is poor, well-structured interventions can overcome disparities in health behaviors. Furthermore, this dissertation demonstrates a non-traditional use of randomized trial data in understanding problematic exposure-outcome relationships. Replication of these analyses should be performed using data from randomized trials of different interventions with emphasis placed on examining patterns of intervention effects across levels of potential effect modifiers rather than on evaluating statistical significance.

Table of Contents

Table of Contents

I. Introduction and Motivation

a. Chronic Disease Burden and Preventability

b. Low Physical Activity and Fruit and Vegetable Consumption as Risk Factors

c. Research Objectives

d. References

II. Literature Review

a. Physical Activity Behavior and Environment

b. Fruit and Vegetable Consumption and Environment

c. Loss to Follow-up

d. Rationale and Significance of Studies

e. References

III. Methods

a. Study Population

b. Assessment of the Environment

c. Statistical Analysis

d. Systematic Review of Fruit and Vegetable Consumption Interventions

e. References

IV. Effects of park access and street structure on randomized trial efficacy

a. Abstract

b. Introduction

c. Methods

d. Results

e. Discussion

f. Acknowledgements

g. Competing Interests

h. Funding

i. References

V. Effects of grocer access on randomized trial efficacy

a. Abstract

b. Introduction

c. Materials and Methods

d. Results

e. Discussion

f. References

g. Analysis Restricted to those Geocoded During the First Round

h. Imputation Analysis

i. Misclassification of grocer businesses by Block Group-Level Demographic Characteristics

j. References

VI. Effects of individual- and block group-level demographics, and built environment on loss to follow-up

a. Results

b. Conclusions

VII. Effect Modification in Randomized Trials of Behavioral Interventions for Increasing Fruit and Vegetable Consumption

a. Abstract

b. Introduction

c. Methods

d. Results

e. Discussion

f. References

VIII. Conclusions

a. Summary of Findings

b. Strengths and Limitations

c. Implications

d. References

IX. Appendices

a. Appendix 1: Health for Life Intake Questionnaire (March 2005)

b. Appendix 2: Get on Track, Stay on Track Outline (March 2005)

c. Appendix 3: Health for Life Counseling Outline (March 2005)

d. Appendix 4: Health for Life Evaluation 3, 6, and 12 months (March 2005)

e. Appendix 5: Power Simulation

f. Appendix 6: Search Strategy for Systematic Review

About this thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
  • English
Research field
Committee Chair / Thesis Advisor
Committee Members
Last modified

Primary PDF

Supplemental Files