Main content area

Mississippi Communities for Healthy Living: Implementing a nutrition intervention effectiveness study in a rural health disparate region

Connell, Carol L., Thomson, Jessica L., Huye, Holly F., Landry, Alicia S., Crook, LaShaundrea B., Yadrick, Kathy
Contemporary Clinical Trials 2015 v.42 pp. 196-203
Whites, behavior change, blood pressure, cross contamination, data collection, experimental design, females, linear models, nutritional adequacy, nutritional intervention, planning, quality of life, rural health, rural population, self-efficacy, social support, sociodemographic characteristics, vegetables, Mississippi
Background: Intervention research in rural, health disparate communities presents unique challenges for study design, implementation, and evaluation. Challenges include: 1) culturally appropriate intervention components; 2) participant recruitment and retention; 3) treatment cross-contamination; 4) intervention delivery and data collection; and 5) potential measurement reactivity. Purpose: The purposes of this paper are to 1) detail the methods of the MCHL study and 2) report baseline demographic characteristics of study participants. The secondary aim is to determine if study participants were engaging in behavior changes after enrollment and prior to intervention initiation. Methods: MCHL was developed using the RE-AIM planning and evaluation framework (reach, effectiveness, adoption, implementation, maintenance). Intervention components were based on Roger’s diffusion of innovation attributes that promote adoption of a new innovation as well as on the psychosocial constructs of social support, self-efficacy and decisional balance. Rolling enrollment data collection was used to acquire sufficient sample size and a second data collection just prior to intervention implementation assessed measurement reactivity effects. Participant outcomes included diet quality, blood pressure, weight status, and quality of life. Cluster stratified assignment to one of two treatment arms was utilized to minimize cross contamination. Generalized linear models were used to compare enrollment measures between the two treatment arms while mixed model linear regression was used to test for changes in diet quality outcomes from enrollment to pre-intervention baseline. Results: There were no significant differences in participant demographic, anthropometric or clinical measures between the two treatment arms at enrollment. With the exception of total vegetables, none of the diet quality indicators were significantly different between enrollment and baseline time points. Conclusions: Conducting nutrition intervention research in a rural health disparate region requires flexibility in adapting the recruitment, retention, and data collection procedures while maintaining a high level of scientific rigor. Negligible research participation effects, such as measurement reactivity, were noted in this population. However, further research is needed to identify methods to successfully recruit and retain Caucasian females to participate in community-based nutrition interventions in this region.