Feasibility of an Adapted Community-Based Lifestyle Intervention to Prevent Cancer in the Rural South: Healthy Living Partnerships to Prevent Cancer (HELP PC)

Abstract

Obesity increases risk of cancer onset and promulgates cancer mortality. Healthy Living Partnerships to Prevent Cancer (HELP PC) is an adapted intensive lifestyle intervention that is facilitated by community health workers (CHWs). The primary objective of this one-arm pilot study was to test the feasibility of evaluating HELP PC in a rural community by assessing participant recruitment, retention, and adherence to the intervention. The secondary objectives of this study were to evaluate the feasibility of collecting study measures and analyze intervention effects to inform future studies. Adults of all races and a BMI ≥ 25 kg/m² who resided in the Dan River Region of Southern Virginia were recruited. Participants received 24 weekly (hour-long) group sessions led by a CHW and two consultations with a registered dietitian (RDN). Seventy-five percent (21/28) of eligible subjects were enrolled (n = 21; mean age = 46 years; 67% African American; 90% female; median BMI = 36.1), and recruitment was completed in 2 weeks. Fifty-two percent (11/21) of participants attended >70% of group sessions (adherence) and 98% of RDN consultations were attended. Eighty-six percent (n=18) of participants completed the 6-month follow-up visit (retention), and showed improvements in moderate physical activity, health literacy, general health, energy, and emotional well-being. Feasibility of HELP PC was established through efficient participant recruitment, modest attendance, high retention, and execution of data collection procedures. Importantly, findings can be applied to advance cancer prevention lifestyle interventions in rural communities.

Introduction

Background

Obesity increases risk of incident cancer and mortality, and obesity-related cancers account for approximately 40% of all cancers diagnosed in the USA, including thyroid, breast, colorectal, and pancreatic cancers [1]. All-cause cancer mortality rates are higher in rural areas than urban areas, which has been attributed to numerous factors including health care workforce shortages, less access to cancer prevention and screening services, lower socioeconomic status, and greater transportation barriers [2]. Moreover, rural residents are more likely to engage in risk behaviors, including tobacco use and physical inactivity, and obesity prevalence is significantly higher in rural communities (34.2 %) than urban areas (28.7 %) [3]. In fact, rural stakeholders report that the 2nd most important health issue facing rural communities is “nutrition and weight” which is only surpassed in priority ranking by “access to quality health services” [4]. Scalable, cost-effective interventions are needed to reach rural populations at high risk of obesity-related cancer.

The Healthy Living Partnerships to Prevent Cancer (HELP PC) lifestyle intervention (Table 1) is based on the Healthy Living Partnerships to Prevent Diabetes (HELP PD) study [5, 6] completed in 2012 which was a weight loss intervention adapted from the Diabetes Prevention Program (DPP) research study and facilitated by community health workers (CHWs). HELP PD demonstrated effectiveness in reducing weight (by 7%), waist circumference, fasting blood glucose (by 4 mg/dL), and insulin resistance in a suburban population, and at significantly lower cost than the DPP [7, 8].

Table 1.

HELP PC intervention

Contact schedule

Intervention objectives

24 weekly (1 h) CHW-led group meetings Held in local cancer resource center and at a worksite Content delivered via DVD series or by community experts Educational content and social support 2 individual RDN consultations Baseline, 6 months

Knowledge Health benefits of weight management and fitness for cancer prevention Nutrition and energy balance basics for cancer prevention Appropriate weight loss and exercise methods, and precautions Diet Reduction of intake by 500–1000 kcal/day Physical activity Gradual progression to 180 min/week of moderate intensity exercise Behavioral Skills Goal setting, self-monitoring, self-reinforcement Stimulus control, cognitive restructuring

Objectives

The primary objective of this pilot study was to test the feasibility of implementing and evaluating an adapted version of the HELP PD intervention for cancer prevention (HELP PC) in a rural community setting—the Dan River Region of Southern Virginia. The following outcomes were assessed: feasibility to recruit and retain participants, adherence to the intervention, and completion of biomarker data collection. The secondary objective of this study was to assess immediate effectiveness of HELP PC in reducing obesity-related cancer risk factors. Results from this one-arm pilot study will inform development of a larger HELP PC trial in similar rural settings.

Methods

Intervention

HELP PC is an intensive lifestyle intervention focused on weight reduction (5–7% over 6 months) through reduced calorie intake (500–1000 kcal/day goal) and increased physical activity (at least 180 min/week goal), and was a modification of the HELP PD diabetes prevention intervention developed for a previous study [5, 6]. The design of HELP PD [6] incorporated major components of the DPP curriculum, including educational content covering key concepts in nutrition, physical activity, and lifestyle change maintenance strategies. However, HELP PD was adapted to be delivered to groups by community health workers (CHWs), in community-based settings (such as recreation centers, churches, libraries.) with individual nutrition counseling and CHW supervision provided by RDNs.

Based on social cognitive theory (SCT), HELP PC targets improvements in self-efficacy, outcome expectations, behavioral capability, self-monitoring, and reinforcement. Positive changes in cognitive-behavioral outcomes are achieved through tailored delivery of social support, health education, and skills training in the following: (1) nutrition and exercise basics, (2) energy balance, (3) healthy eating, (4) goal setting, and (5) problem-solving using the continuous care problem-solving model [6].

Two RDNs employed by the Danville Regional Medical Center provided two personal consultations with participants and were trained to monitor and support a CHW who facilitated 24 weekly group sessions (Table 2). To ensure fidelity to the curriculum, educational content was delivered through a DVD series developed by the research team for HELP PD and followed with a guided group discussion facilitated by the CHW. All intervention materials adhere to a sixth-grade literacy level. Prior to initiation of this HELP PC pilot study, the CHW lesson plans, participant workbooks, and DVD series underwent minor adaptions to focus on general weight loss strategies for cancer prevention (versus weight loss for diabetes prevention), and planned discussions held during group sessions emphasized the relationship between obesity and cancer risk. In addition, the presenter who was selected to narrate the HELP PC DVD series was a regionally recognized TV news anchor of African American race. Experts from the local rural community (such as representatives from the YMCA, local grocery store, and footwear experts) also participated in HELP PC group sessions to introduce participants to available resources for healthy living. Makeup sessions were offered by phone or in-person.

Table 2.

HELP PC weekly session topics

Session #	Session title
1	Welcome to HELP PC
2	Nutrition 101
3	Physical Activity 101
4	Footwear
5	Troubleshooting
6	Calorie Balance
7	Mindfulness
8	Portion Sizes
9	Troubleshooting
10	Community Exercise/Nutrition Resources
11	Problem Solving
12	Physical Activity Hands On
13	Troubleshooting
14	Emotions and You
15	Healthy Eating
16	Stretching/Injury Prevention/Strength Training
17	Troubleshooting
18	More about Healthy Eating
19	Food Shopping/Eating Out
20	Creating an Environment for Success
21	Troubleshooting
22	Weight loss Maintenance
23	Preparation for Independence
24	Transition

The CHW underwent 20 h of training led by the research team. CHW competencies included HELP PC curriculum and materials; metabolic risk factors for obesity-related cancers; social support delivery; group facilitation; participant coaching and goal reinforcement; data collection; and attendance tracking. Consistent with social cognitive theory–based behavior change strategies, the CHW personally modeled HELP PC protective health behaviors including regular exercise, maintaining a healthy weight, and healthy eating. The CHW’s race and ethnicity matched the majority of intervention participants [9] and he was a member of the targeted rural community.

Setting

Participants resided in the Dan River Region of Southern Virginia, and were recruited from a local community center and a worksite. Delivery of the intervention was conducted at the worksite and a local cancer resource center in Danville. Danville is a small-medium city with a population of 40,444 that serves as the economic hub for a large region of adjoining rural counties in southern Virginia and North Carolina [10]. The city of Danville is comprised of 51.5% African American, 44.9% White, 4.7% Hispanic, 1.4% Asian, 0.5% American Indian or Alaskan Native, and 0.1% Native Hawaiian or Pacific islander residents. By statewide comparison, Danville has a markedly greater percentage of African American residents (51.5% vs. 19.9% statewide) and a smaller percentage of White residents (44.9% vs. 69.4% statewide). In addition, Danville residents are less likely to have health insurance, are much less likely to have completed a high school bachelor’s degree, and are twice as likely to be living in poverty than Virginia residents statewide [11].

Participants

It was anticipated that enrollment of a target sample of 21 participants could be accomplished in 3 months, averaging 1–2 participants per week over the course of 12 weeks. English-speaking adults of all ages and races with overweight or obesity (BMI ≥ 25) were recruited. Candidates were excluded based on the following criteria: current or planned pregnancy (within 1 year), uncontrolled hypertension, recent history of cardiovascular disease in the past 6 months, cancer or other serious chronic disease in past 5 years, use of diabetes medication or insulin, or current enrollment in a supervised weight loss program. Participants were approached for participation by the CHW at a local community center, and flyers and study-related materials were placed in community centers and health facilities in the Danville, Virginia area. Interested potential participants were screened via telephone by a data collector at Wake Forest School of Medicine.

Measures

Demographics (self-reported age, gender, race, and ethnicity) and study eligibility criteria were collected by telephone by research personnel at Wake Forest School of Medicine. Anthropometric measurements (weight, height, and waist circumference) were collected by research personnel and graduate students from Virginia Polytechnic Institute and State University (Virginia Tech) in a local community center. Waist circumference was measured at the smallest horizontal circumference in the area between the ribs and iliac crest. Systolic and diastolic blood pressure were also measured twice and averaged, at baseline and follow-up following JNC-7 guidelines for blood pressure measurement. Fasting blood samples were collected at baseline and 6 months for fasting (8 h) plasma glucose, insulin, and cholesterol (total, LDL, HDL, triglycerides) at a local LabCorp facility in Danville, VA. Participants were instructed to fast for at least 8 h prior to this visit. Study personnel also collected self-reported information on prescription medications at both time points.

Health-related quality of life (HRQL) was measured at baseline and 6 months using the Short Form 36 (SF-36) [12], which is comprised of 8 scales measuring the following domains: general health, physical function, physical role limitations, emotional role limitations, energy level, social function, emotional well-being, and pain. Each scale is scored within a range of 0 to 100, with higher scores representing better self-reported health status.

Physical activity was measured using the International Physical Activity Questionnaire (IPAQ) short form [13], which consists of a 7-item index measuring weekly time spent engaging in light, moderate, and vigorous physical activity. Time engaged in physical activity is multiplied by an assigned metabolic equivalent (“MET”) for each intensity level and expressed as MET-minutes per week. Total physical activity includes vigorous activity, moderate activity, brisk walking, and regular walking. In addition, participants were asked to monitor their daily caloric intake and physical activity (duration and type) using a food and exercise diary that was provided by the research team and review their diary entries each week with the CHW for feedback. The diaries were also used by the RDNs to provide advice for healthier food choices and caloric restriction during individual counseling visits.

Health literacy was measured using a validated and reliable screening tool, the Newest Vital Sign (NVS) assessment [14]. Participants were provided a nutrition label for ice cream and research staff asked seven questions requiring their interpretation of the label. For example, participants were asked, “If you eat the entire container, how many calories will you eat?” (correct answer: 1,000 calories). One point is earned for each correct answer, for a total possible score of 7.

Participant satisfaction with the intervention and with their own performance was also collected at follow-up, using an anonymous survey that was not linked to study data. This survey consisted of 10 items measuring satisfaction on a 7-point Likert scale (−3 = very dissatisfied, +3 = very satisfied) and 4 open-ended questions.

Statistical Analysis

The target sample size for this study (n = 21) was determined sufficient to estimate feasibility as well as mean changes in continuous outcomes, and inform power analyses for subsequent studies. Descriptive analyses were performed to illustrate the feasibility of conducting a larger HELP PC trial, including participant characteristics, recruitment, adherence, and retention to the pilot intervention. A Fisher’s exact test and Mann-Whitney U test were used to determine differences in participant characteristics between those who adhered to the intervention (completed ≥ 70% of group sessions) and those who did not, and between those who completed follow-up (retention) and those who did not. A Wilcoxon signed-rank rest was used to assess pre-post differences in all measurements, including anthropometric, clinical, and behavioral measures described above. Estimated effects sizes were calculated by dividing the Wilcoxon statistic by the total rank sum. IBM SPSS Statistics (Armonk, NY) version 26 was used for all analyses.

Results

Participant Recruitment

A total of 33 interested candidates (Figure 1) contacted the study team via telephone to inquire about the study and of those, 28 were screened. All 28 were deemed eligible to participate in the study and were scheduled for an appointment to complete informed consent, enrollment, and baseline measurements. Of the 28 eligible participants, 21 (75%) attended their scheduled appointment and were enrolled. Baseline and follow-up anthropometric, vital signs, and study questionnaires were collected by research personnel and graduate students from Virginia Tech. Phlebotomy was performed at a local Danville LabCorp facility by trained phlebotomy staff.

Fig. 1.

Pilot flow diagram

Recruitment was completed in 2 weeks, much sooner than expected. Most candidates were directly approached by the HELP PC CHW at the local community center or were employed at the worksite that offered the opportunity to participate in the program; four participants (19%) reported that they heard about the study through a friend.

Baseline Participant Characteristics

Mean age was 46 years (Table 3). Most participants were African American (67%) and female (90%). Median BMI at baseline was 36.1 (Table 4); all participants had overweight or obese status, and most (81%) had a BMI over 30 (class 1 obesity 24%; class 2 obesity 29%; class 3 obesity 29%). Median waist circumference was 104.4 cm. Median fasting plasma glucose was 95.0 mg/dl; five participants met criteria for prediabetes (100–125 mg/dL). Median fasting insulin was 11.2 μU/mL; three participants had a fasting insulin over 25 μU/mL. Median total cholesterol was 181.0 mg/dL; median HDL cholesterol was 55.0 mg/dL; median LDL cholesterol was 111.0 mg/dL. Median triglycerides were 79.0 mg/dL. Among HRQL domains, self-reported energy levels were markedly low (47.5 out of 100), followed by self-reported general health (62.5 out of 100). Both moderate and vigorous baseline physical activity levels (reported as MET-minutes per week) were very low (moderate, 240 METs; vigorous, 60 METs). Baseline health literacy was adequate (4.5).

Table 3.

Baseline participant characteristics (N=21), adherence, and retention

Characteristic	Total	Adherence		Retention
	n (%)	n (%)	p value	n (%)	p valuea
Gender			1.0		1.0
Male	2 (10)	1 (50)		2 (100)
Female	19 (90)	10 (53)		16 (84)
Race			0.66		0.25
Black	14 (67)	8 (57)		13 (93)
White	7 (33)	3 (43)		5 (71)
Age, mean (SD)	45.7 (11.6)	-	0.02*	-	0.08

^aResults of Fisher’s exact test (gender, race) and Mann-Whitney U test (age)

^*Statistically significant

Table 4.

Intervention measurements, pre-post (N=18)

Measurement	Baseline	Follow-up	r
BMI	36.3 (30.4; 45.4)	35.8 (29.8; 47.0)	−0.01
Weight (pounds)	216.7 (172.8; 261.0)	218.5 (177.3; 272.1)	0.07
Waist circumference (cm)	103.8 (93.5; 118.0)	103.0 (96.5; 117.0)	0.10
Glucose (mg/dL)	95.0 (89.0; 104.0)	96.0 (90.0; 100.0)	0.10
Insulin (μU/mL)	11.4 (9.1; 21.1)	11.3 (7.2; 22.2)	0.05
Systolic BP (mmHg)	118.5 (112.0; 134.0)	122.5 (115.0; 137.0)	0.27
Diastolic BP (mmHg)	77.5 (69.0; 83.0)	81.0 (74.0; 87.0)	0.27
Total cholesterol (mg/dL)	186.0 (156.0; 199.0)	187.0 (173.0; 210.0)	0.14
LDL	112.0 (92.0; 128.0)	120.0 (110.0; 127.0)	0.22
HDL	57.0 (48.0; 67.0)	51.0 (43.0; 68.0)	−0.08
Triglycerides (mg/dL)	79.0 (51.0; 91.0)	71.0 (61.0; 105.0)	−0.01
Health status (SF-36) (range 0–100)
General health	62.5 (50.0; 75.0)	70.0 (55.0; 80.0)	0.2
Physical functioning	95.0 (90.0; 100.0)	97.5 (80.0; 100.0)	0.06
Role limitations—physical	75.0 (50.0; 100.0)	100.0 (75.0; 100.0)	0.10
Role limitations—emotional	83.3 (33.3; 100.0)	100.0 (33.3; 100.0)	0.03
Energy	47.5 (20.0; 65.0)	55.0 (40.0; 75.0)	0.30
Emotional well-being	72.0 (64.0; 84.0)	76.0 (72.0; 88.0)	0.15
Social functioning	75.0 (62.5; 100.0)	75.0 (62.5; 100.0)	0.00
Pain	78.8 (67.5; 90.0)	80.0 (67.5; 90.0)	0.02
Physical activity (MET-minutes/week)
Vigorous	60.0 (0.0; 720.0)	60.0 (0.0; 360.0)	−0.05
Moderate	240.0 (80.0; 420.0)	360.0 (240.0; 780.0)	0.22
Total	360.0 (200.0; 1260.0)	680.0 (360.0; 1080.0)	0.05
Health literacy (NVS)	4.5 (3.0; 6.0)	6.0 (5.0; 6.0)	0.20

Values presented are median (25th percentile, 75th percentile)

Adherence

More than half (59%, mean = 13 ± 6.4 sessions) of scheduled group sessions were attended, and 52% of participants completed at least 70% of group sessions. Older participants adhered to the intervention more than younger participants (Mann-Whitney U = 22.0, p = .02). There were no significant differences in race, gender, BMI, weight, or health status among those who adhered to the intervention and those who did not (Table 3).

One participant never attended a single group session and the CHW’s multiple attempts to contact the participant were unsuccessful. Two other participants stopped attending group sessions: one withdrew after the initial session due to work and school commitments, and the second participant withdrew after the 12th session due to work-related scheduling conflicts. The remaining 18 participants stated work commitments, family emergencies, funeral, and illness as reasons for absences.

Of the two dietitian consultations offered, all participants completed the first consultation which was a part of the orientation meeting, and 20 participants (95%) completed the second RDN consultation at 3 months. Therefore, total attendance at RDN consultations was very high (98%). The food and exercise diaries were not systematically completed by participants or collected by the CHW.

Retention

Retention was high, as 86% (18/21) of participants completed the follow-up visit at 6 months (Figure 1). Two participants did not complete follow-up due to work demands; one participant was unable to be reached. There were no significant differences in participant characteristics among those who completed follow-up compared to those who did not; however, difference in age approached significance (Mann-Whitney U = 9.0, p = 0.08) (Table 3).

Intervention Effects

Among participants who completed follow-up (n=18), a Wilcoxon signed-rank test revealed improvements (with small and medium effects) in the following variables: energy (SF-36), (r = .30); general health (SF-36), (r = .20); health literacy, (r = 0.20); moderate physical activity, (r = 0.22); and emotional well-being (SF-36), (r = .15) (Table 4).

Participant Satisfaction

Participants (n = 18) reported that they were very satisfied with the CHW (2.8 ± 0.4), the group sessions (2.5 ± 0.9), and the information provided by guest speakers from the community (2.5 ± 0.8), and somewhat satisfied with the Registered Dietitian’s services (2.2 ± 0.9) and the overall schedule and location of the intervention (2.4 ± 1.2). Four participants (n = 18) cited scheduling conflicts related to their work or other time commitments. In addition, participants reported that they were least satisfied with their ability to attend group sessions (0.8 ± 2.3), meeting their calorie goals (0.0 ± 2.1), and their weight loss (−0.1 ± 2.1). However, positive changes included “better self-image, more confident and motivated to lose weight,” “my awareness of weight [related] to cancer,” “reading labels and contents more carefully,” and “talked with my PCP about weight/stress reduction.”

Discussion

Healthy Living Partnerships to Prevent Cancer (HELP PC) is a community- and group-based adaptation of the DPP lifestyle intervention, managed by RDNs who are already employed by a local health facility and CHWs who serve as intervention facilitators and vital community-clinical linkages to existing community preventive resources for weight loss and cancer prevention.

This pilot study demonstrated the ability to recruit (75%) and retain (86%) participants living in a rural Virginian community. Moreover, direct recruitment efforts by the CHW and study investigators accelerated this study’s participant accrual rate by 84%, completing recruitment in 2 weeks versus 12 weeks.

Recruitment in rural populations is typically challenging, and participation in research studies among rural residents is historically lower than urban residents [15]. Evidence suggests that cultural differences in rural communities (versus urban) may impact participation in clinical trials. For example, rural residents tend to value self-sufficiency and independence, and believe that work results in good health [16]. Indeed, work-related priorities were cited by participants in our study, and even though group sessions were offered at the worksite, attendance was modest (52% attendance ≥ 70% of sessions), with older participants demonstrating higher adherence. These findings are consistent with previous worksite interventions that reported greater adherence among older participants, and cited the importance of having strong support for the program among worksite leadership [17]. Indeed, while participants were very satisfied with the intervention, they reported low satisfaction with their own attendance. Future iterations of HELP PC should address work-related scheduling challenges that may include adding more times for HELP PC sessions and strengthening intervention support from worksite leadership. Moreover, framing desirable outcomes such as increased work productivity through lifestyle intervention may help address the perception among rural residents (and worksite leadership) that work is more important than health needs, and thereby improve adherence [16].

Furthermore, this study demonstrated a unique approach to cancer prevention—by using a modified translation of the DPP lifestyle intervention. To our knowledge, no other interventions of this nature have been reported. Likewise, our study demonstrated the feasibility of the HELP PC intervention, which was previously developed in a highly educated suburban population, and reconstituted for cancer prevention in a comparatively disadvantaged, rural, minority population. These findings may help inform future innovative translations in similarly under-resourced settings characterized by health care workforce shortages, including the rural south where the rural-urban gap in cancer mortality continues to grow and preventive and screening services are lacking. The HELP PC pilot employed several recommended strategies that are well-suited for rural areas including (1) collaborative partnerships with rural hospitals and health centers; (2) maximizing existing community resources for program facilitation and delivery, including recreation centers, community experts, and CHWs; and (3) mobilizing rural providers (RDNs) for cancer prevention services [18].

Limitations of this study included its small sample size, single-arm design, and lower attendance rate which may have reduced immediate intervention effects on measured outcomes. Future trials may also explore strategies to overcome accessibility barriers that exist in rural areas including the provision of transportation alternatives and telehealth modalities [16]. In addition, future trials should address challenges related to participant completion of the HELP PC food and exercise diary and include a more formal assessment of dietary intake appropriate for a larger sample size.

Strengths of this study included the use of recently identified strategies to recruit and retain rural participants, including the provision of social support and establishing trust between participants and research team members through a local peer health leader (CHW). Furthermore, racial and ethnic minorities are historically under-represented in clinical trials [15]. Based on population estimates of Danville, VA, we expected to recruit at least 50% African American participants. Actual enrollment was 67% African American, providing a representative sample of participants in this study.

Effective, scalable, and sustainable cancer prevention interventions are needed to address rural cancer disparities—where limited preventive care resources and provider shortages exacerbate the rural-urban cancer mortality gap. Despite this study’s limitations, HELP PC participant accrual and retention rates were excellent, and it is evident that people living in this rural community were interested in participating in a cancer prevention lifestyle intervention. A HELP PC effectiveness study in a larger, higher powered, sample is warranted.

Availability of Data and Material

The study was pre-registered at clinicaltrials.gov and can be found at https://clinicaltrials.gov/ct2/show/NCT02960542. The analysis plan was not formally pre-registered. De-identified data from this study are not available in a public archive. Analytic code used to conduct the analyses presented in this study is not available in a public archive. They may be available by emailing the corresponding author. Materials used to conduct the study are not publically available.