A systematic review of lifestyle interventions for chronic diseases in rural communities

Background Rural Americans suffer disproportionately from lifestyle-related chronic diseases (e.g., obesity, diabetes, hypertension, cardiovascular disease, and breast cancer). Interventions that consider the distinctive characteristics of rural communities (e.g., access to healthcare, income, and education) are needed. As an initial step in planning future research, we completed a systematic review of dietary intake and physical activity interventions targeting rural populations. Methods Manuscripts focused on dietary intake and physical activity and published through March 15, 2016, were identified by use of PubMed and CINAHL databases and MeSH terms and keyword searches. Results A total of 18 studies met the inclusion criteria. Six involved randomized controlled trials; 7 used quasi-experimental designs; 4 had a pre-/post-design; and 1 was an observational study. Eight studies were multi-site (or multi-county), and 3 focused on churches. Primary emphasis by racial/ethnic group included: African Americans (6); Whites (2); Hispanics (3); and two or more groups (7). Most studies (17) sampled adults; one included children. Two studies targeted families. Conclusions Additional lifestyle intervention research is needed to identify effective approaches promoting healthy diet and exercise and chronic disease prevention in rural communities. Studies that include rigorous designs, adequate sample sizes, and generalizable results are needed to overcome the limitations of published studies.


INTRODUCTION
related to dietary intake and physical activity targeting rural populations in the US. The goals were to examine the effectiveness of lifestyle interventions in rural communities and to determine salient features for replication in future studies.

METHODS
The present review is based upon bibliographic searches of PubMed and CINAHL with relevant search terms. Articles published in English through March 15, 2016, were identified using the following MeSH search terms and Boolean algebra commands: (((dietary intake) or (diet) or (nutrition) AND (physical activity) or (exercise)) AND (rural)). Although the search criteria did not specify a begin date, the earliest article that met the search criteria was published in 2004. The searches were not limited to words appearing in the titles of articles. Information obtained from bibliographic searches (title and topic of article, information in abstract, geographic locality of a study, and key words) was used to determine whether to retain each article. In addition, reports included in Cochrane reviews (http:// community.cochrane.org/cochrane-reviews) were identified. Only completed studies were included; study protocols were eliminated ( Figure 2); 154 citations were found.
After screening the abstracts or full texts of these articles, 50 studies of rural populations were identified, all of which were conducted in the US. The inclusion criteria were completed studies of interventions focusing on dietary intake and/or physical activity and performed in rural areas in the US. Thirty-two reports that did not meet the inclusion criteria were not considered further. The two eligible studies identified in the CINAHL search overlapped with those identified in the PubMed search.

RESULTS
A total of 18 studies met the search criteria (Table 1). Six included a randomized controlled approach; 7 used quasi-experimental designs; 4 had a pre-/post-design; and 1 was an observational study. Eight studies were multisite or multi-county, and 3 focused on churches or other institutions within rural communities. Primary emphasis by racial/ethnic group included: African Americans (6), Whites (2), Hispanics (3), and two or more groups (7). Most studies (17) focused on adults; 1 included children. Two studies targeted families.  conducted a quasi-experimental study with 35 White obese breast cancer survivors aged 46-74 years from three cancer centers in rural Kansas. The 6-month, groupbased, weight control intervention incorporated self-regulation skills and social support to enhance changes in diet and physical activity. There were significant post-intervention changes for weight (>10% loss), fruit and vegetable consumption (+3.7 ± 4.3 servings/day), and physical activity (+1235 ± 832 kcal/week), and reductions in two biochemical mediators of breast cancer risk, fasting insulin (−16.7%) and leptin (−37.1%).
greater reductions in BMI relative to men. Participants with the most NCM contacts attended more DSME sessions, and higher attendance resulted in greater reductions in HbA 1c levels. Carter et al. (2015) engaged 55 African American men and women in a year-long intervention in two rural Alabama counties (Bullock and Macon). For those in Macon County, physical activity included membership in Curves, floor exercise, or walking; those in Bullock County engaged in aerobics, dance, the gym, or walking. Individuals doing floor exercise lost the most weight (22.4 lb or 11.18% change) followed by those walking (6.1 lb or 3.40% change). For all physical activities, systolic and diastolic blood pressure decreased respectively). Fahs et al. (2013) compared a stage-matched nursing and community intervention (SMN +CI) to CI alone in a 14-month, multisite, randomized control trial of 117 rural White, Black, and Hispanic women aged 35-65 years in rural counties in New York and Virginia. Significant increases in fruit and vegetable consumption and reduced diastolic blood pressure were noted in the SMN+CI cohort. The CI group had a significant reduction in total cholesterol.
In rural central Nebraska, Hageman et al. (2014) conducted a 12-month randomized controlled trial to reduce blood pressure among 289 women (primarily White), ages 40-69 with hypertension. Participants were randomized to standard advice, web-based intervention, or print intervention. For the web-based and print intervention cohorts, waist circumference, % calories from fat, saturated fat, servings of fruits and vegetables, and low fat dairy improved significantly. Improvements were observed in web-based vs. standard advice groups in systolic blood pressure (p=0.048) and estimated VO 2 max (p=0.037).
Focusing on diabetes self-management behaviors, Hu et al. (2014) completed a quasiexperimental, family-based study of 36 Hispanics (mostly women) and 37 members of their adult families. The 8-week pilot study, conducted in rural central North Carolina, consisted of two family sessions and eight weekly group sessions. Session content included education on diabetes self-management, exercise, food, and eating healthy. Among study participants, HbA 1c dropped by a mean of 0.41% from baseline to 1-month post-intervention.
In two rural counties in southwest Georgia, Kegler et al. (2012) conducted a coach-based, quasi-experimental intervention to improve food quality and physical activity in 90 households. Participating in the 6-week study were African American and White adults ages 40-70 years and their household members 18 years of age or older. Intervention households reported increased exercise relative to comparison households. Kim et al. (2008) conducted a study of a faith-based weight loss intervention. With an intervention group and a delayed-intervention control group, a quasi-experimental design was used. The 73 participants (71% female, mean age 54.1 years, 100% African American) were from rural churches in North Carolina. Small groups led by trained community members met weekly for 8 weeks. The community members emphasized physical activity, healthy nutrition, and the connection of faith to health. The mean weight loss in the intervention group was 3.60 lb relative to 0.59 lb in the control group. Landry et al. (2015) completed a 6-month, community-based, pre-post trial of an intervention consisting of motivational enhancement, social support, pedometer diary selfmonitoring, and educational sessions. The participants were 269 adults (94% African American, 85% female, mean age 44 years) in Hattiesburg, Mississippi. The outcome measures were steps per day, fitness, dietary intake, and psychosocial construct measures. For the physical activity and dietary outcome variables, there were temporal changes only for steps per day and sugar intake. Sugar intake decreased by about 3 teaspoons, and physical activity increased by approximately 2,010 steps per day.
In a multi-site, cardiovascular disease prevention study, Lilly et al. (2014) investigated barriers to lifestyle change among 81 white, Hispanic, and African American women in Colorado, North Carolina, and West Virginia. Most participants (72%) reported significant post-intervention improvements in problem-solving skills (p<0.001), perceived stress (p<0.05), and maintenance of or increases in fruit and vegetable intake and physical activity.

Mayer-Davis et al. (2004) conducted a 12-month randomized clinical trial of 152 African
American and White diabetic men and women in rural South Carolina. The study evaluated the effectiveness of a Diabetes Prevention Program-type intervention (intensive intervention), reimbursable-lifestyle intervention, and usual care (controls) on weight loss and HbA 1c . Relative to controls, intensive-intervention participants had greater weight loss than controls (−2.6 kg vs. −0.4 kg, p<0.01). HbA 1c was reduced among all participants (p<0.05) but was not different between cohorts.
In three rural counties in South Carolina, Parker et al. (2010) conducted a church-based weight loss intervention among African American women. The study was developed with community involvement. The 35 participants, between the ages of 25 and 64 years, were not pregnant or breast-feeding. Two 10-week interventions (spiritually-based and nonspiritually-based) were pilot tested using a pre-post design. Physical activity was assessed using the Yale Physical Activity Survey. Both interventions led to significant reductions in BMI, but the spiritually-based intervention (z = −1.97, P<0.01) led to greater reductions. For the spiritually-based group, significant improvements were found for physical activity (z = −2.74, P<0.01).
In three rural North Carolina counties, Ries et al. (2014) conducted a project with a quasiexperimental design. The participants were 485 low-income, predominately minority women (63% African American) with a mean age of 47.5 years. The curriculum for the bi-weekly group meetings, held over a 6-month period, addressed physical activity, healthy eating, weight control, stress management, education, and job skills. For both African Americans (P<0.05) and whites (P<0.0001), intervention participants were more likely than comparison participants to move from contemplation to action/maintenance in regard to the goal of increasing physical activity. For all participants, progression in stages of change mediated the intervention effect on physical activity but not fruit and vegetable intake. Intervention group participants engaged in more minutes of physical activity per week (138 minutes) than comparison participants (86 minutes, P≤0.05). Robles et al. (2014) conducted a 12-week prospective study of 33 children aged 8-11 years in Bailey County, Texas. The purpose was to assess the impact of a pharmacy-directed pilot study on dietary intake and physical activity during out-of-school time. Post-intervention changes were noted in BMI (−0.30, p<0.0001), waist circumference (−0.47, p<0.001), decreased consumption of fried/sweet foods, and increased exercise.
In Alabama's "Black Belt," Scarinci et al. (2014) conducted a community-based, clusterrandomized trial comparing two interventions: 1) promotion of physical activity and healthy eating (healthy lifestyle arm), and 2) promotion of breast and cervical cancer screening. Participants were 565 African American women of ages 45-65 years in six rural counties in Alabama. At the 12-month follow-up, participants in the healthy lifestyle arm showed significant positive changes (increased physical activity, increased fruit/vegetable intake, and decreased consumption of fried food). At 24 months, these positive changes were maintained for healthy eating behaviors but not for physical activity.
In a quasi-experimental, 6-month intervention related to diet and physical activity, Tussing-Humphreys et al. (2013) assigned eight churches to intervention or control. Conducted in the Lower Mississippi Delta Region, the study included 403 African American men and women with a mean age of 47 years. Adapted from the Body and Soul program (Resnicow et al., 2004), Delta Body and Soul included peer counseling, a focus on regional cuisine, a didactic physical activity session, and self-directed physical activity. In both study arms, there were significant increases for consumption of total fruits and vegetables (0.3[1.8] and 0.2[1.1]), aerobic physical activity (22%), and strength/flexibility (24%).
In the "Wellness for Women" clinical trial, Walker et al. (2009) randomly assigned 225 women from two similar rural areas to cohorts receiving a computer-tailored newsletter or a generic newsletter. The 12-month intervention and 12-month follow-up examined behavioral markers of activity and eating. Relative to the generic cohort, at 6 months and at 12 months, the 'tailored' group achieved greater strength (p=0.008 and p=0.002) and consumed a lower percent of calories from saturated fat (p=0.028 and p<0.001), respectively. Zoellner et al. (2007) conducted a quasi-experimental study to evaluate a 6-month intervention focused on promoting physical activity and health through walking teams led by coaches, with self-monitoring and monthly 1-hour educational sessions. The participants were 83 rural residents in Hollandale, Mississippi (99% African American, 97% women). There were improvements in waist circumference (−1.4 inches), systolic blood pressure (−4.3 mmHg), and HDL-cholesterol (+7.9 mg/dL) (p<0.001). Self-reported walking per day was 44.8 (SD±52.2) minutes at enrollment and 65.9 (SD±89.7) minutes at 6 months (P=0.154).

DISCUSSION
In the US, rural communities are a large, medically underserved group. Although rural populations in the US have higher rates of chronic diseases than non-rural areas Patterson et al., 2004;Eberhardt et al., 2001), there are few interventions that address lifestyle risk factors among rural populations. We reviewed eighteen interventions targeting lifestyle that were implemented in rural areas in the US. Most of the interventions found, among study participants, significant positive changes in measurable outcomes, such as weight loss, increase in levels of physical activity, and consumption of fruits and vegetables and improvements in biochemical mediators. Only six of the studies used a randomized controlled approach, and only one, by Kegler et al. (2012), was conducted in rural Georgia.
Physical activity and a healthy diet continue to gain recognition as lifestyle interventions for use in primary and secondary prevention (Durstine et al., 2013;Warburton et al., 2006;Willett, 2002). Finding effective methods to reach and influence behavior change in residents of rural communities, who are at high risk for developing chronic diseases, is a public health challenge. In rural America, sociocultural and personal barriers, including the stress of living in poverty, low health literacy, and lack of experience and skill in accessing healthcare information services, remain as obstacles (Brown et al. 2011).
Most of the studies covered in this review employed strategies that were mindful of the socioeconomic barriers to healthcare delivery encountered by rural dwellers. The strategies included community-based participatory research (CBPR), distance-delivery methods (webbased, telephone, and mail), and case management.
CBPR, a framework through which evidence-based interventions (e.g., randomized trials) are developed and implemented in the context of community engagement, offers a partnership approach to research that equitably involves community members, organizational representatives, and researchers in all aspects of the research process (Scarinci et al., 2014). It enhances cultural appropriateness and encourages the building of trust between researchers and community members, thereby facilitating recruitment of participants. The weight-loss intervention by  examined the effect of a group-based, weight control intervention delivered through conference call technology to obese breast cancer survivors living in remote rural locations. The intervention included weekly group phone sessions and a reduced-calorie diet incorporating prepackaged entrees and shakes. Physical activity gradually increased to 225 min/week of moderate intensity exercise. There were also significant changes for weight, diet, physical activity, serum biomarkers, and quality of life. Hageman et al. (2014) also utilized the distance-delivery method. Women in groups who had the intervention delivered by the web or print-mailed, improved more than the group receiving standard advice in regard to waist circumference, daily calories from fat, and daily servings of fruit and vegetables. Lifestyle modification interventions that incorporate web-based components offer advantages of providing tailored messages at low cost in order to reach a large audience across great distances, with convenience for the users. Case management is a strategy to coordinate healthcare services and provide more consistent levels of health care access. Brown et al. (2011) studied changes in diet and physical activity between an experimental group that was administered a diabetes self-management education (DSME) with access to a nurse case manager (NCM), and another group that had DSME only. Although there was non-significant improvement in diet and physical activity between the two groups, the number of NCM contacts was proportional to DSME attendance. Case managers, who help patients locate and manage resources, are advocates within the healthcare system who enhance communication among healthcare providers, patients, and their families (Brown et al. 2011).
The reviewed studies had various limitations. Most (11) had small sample sizes and uncertain generalizability; 5 lacked a control group; and 5 with control groups lacked randomization. Studies by Kegler et al. (2012), Lilly et al. (2014), and Robles et al. (2014) had short intervention periods. In some of the studies, self-reporting of dietary intake and physical activity were limitations.
For rural America and for Georgia in particular, there is paucity of evidence-based studies about lifestyle modification interventions. To achieve health equity between rural and urban communities, increased funding for research activities targeting lifestyle interventions in rural areas is needed. The promotion of more tailored interventions that increase physical activity and consumption of healthy diets among rural residents and that address the limitations of existing studies, is warranted.

CONCLUSIONS
Relative to urban residents, rural residents are disadvantaged in terms of socioeconomic determinants of health, access to and availability of healthcare resources, and facilities that reduce lifestyle risk for chronic diseases. Consequently, among rural communities, the prevalence of chronic diseases is high. Although the few studies on lifestyle interventions in rural areas have generally found favorable outcomes, studies that include rigorous designs and adequate sample sizes and produce generalizable results are needed.   Table 1 Dietary intake and physical activity interventions in rural populations

Study Design Sample Results Limitatios
Befort et al.
Self-reported physical activity outcomes; possible contamination across cohorts Hu et al. The intervention resulted in significant improvement in problemsolving skills (P < 0.001) and perceived stress (P < 0.05). Diet, physical activity, and weight remained stable, although 72% of individuals reported maintenance or increase in daily fruit and vegetable intake, and 67% reported maintenance or increase in daily physical activity.