Exercise in African American and White Colorectal Cancer Survivors: A Mixed Methods Approach

Abstract

Background

African-American (AA) colorectal cancer (CRC) survivors tend to be more obese and less physically active compared to white survivors.

Purpose/Objective

To test the feasibility of an aerobic exercise program as well as explore perceptions about supervised exercise among AA CRC survivors.

Methods

A prospective supervised exercise intervention performed on a cycle ergometer 2 days/week for 12 weeks. Peak (VO_2peak) and sub-maximal exercise (6MWT) along with questionnaires (SF-36, Memorial Sloan Kettering Cancer Center Bowel Function Instrument (BFI), Functional Assessment of Cancer Therapy Scale–Colorectal (FACT-C) and Fatigue (FACIT-F), Brief Symptom Inventory (BSI). A second group of survivors participated in an interview evaluating perceptions regarding exercise.

Design

Prospective case series and qualitative interview.

Setting

Research university and academic medical center.

Patients

African American and white colorectal cancer survivors.

Results

Quantitative: A total of 237 letters were mailed to CRC survivors (112 white, 126 AAs). From the letters, 25 white and 15 AAs expressed interest; only five white (4.5%) and four AAs (3.2%) enrolled. Two AAs and five white survivors (7/9) finished the program. There was an improvement in peak exercise (p=0.011) and quality of life (QoL) (SF-36 total, p=0.035) post-training. Qualitative: 30 CRC survivors (12 AA and 18 white) participated in qualitative interviews and selected co-morbidity, motivation and location as primary barriers to exercise.

Limitations

Small sample size.

Conclusions

Recruiting CRC survivors (regardless of race) into an exercise program is challenging, however, there are exercise and QoL benefits associated with participation. Barriers to exercise are similar between AA and white CRC survivors.

INTRODUCTION

Colorectal cancer (CRC) is the third most common occurring malignancy and third leading cause of cancer-related mortality in both men and women in the United States¹. There are known differences in CRC rates based on race, with African Americans (AA) having a ~23% higher incidence and a ~40% higher mortality rate compared to whites². Moreover, AA men and women tend to be more obese and less physically active compared to their white counterparts with AA women even more physically inactive then AA men³. Leading a physically active lifestyle represents one modifiable risk factor that can positively influence cancer outcomes⁴. In cancer survivors, meeting exercise guidelines is associated with improved survival and a reduced risk of recurrence⁴. For example, men and women living a physically active lifestyle have a 30-40% reduction in the risk of developing colon cancer compared to inactive persons⁴. Whereas, failing to meet guidelines will contribute to an increase in waist circumference and body mass index (BMI), both risk factors associated with CRC risk and metastasis⁵.

It is estimated that less than one-third of CRC survivors and even fewer AA CRC survivors meet recommended exercise guidelines⁶. In spite of this, few exercise interventions have included AAs in their study design or the results are not stratified by race. To the best of our knowledge, only one cross-sectional study has reported improved health-related outcomes among AA CRC survivors who are more physically active⁷. In addition, only one exercise intervention among CRC survivors has described the ethnicity of its participants; 98% were white⁸. The rationale for the current feasibility study is that AAs tend to be at higher risk for CRC and are underrepresented in many exercise trials. We suggest it is important to include diverse patients into supervised exercise trials. There are known barriers to participating in exercise research in minorities and AA’s, such as knowledge of available clinical trials⁹, mistrust in research¹⁰, or being asked to participate in a study¹¹. However, it is unknown what specific barriers and/or facilitators to participation in a supervised exercise program exist in AA CRC survivors.

The purpose of the current study was to explore the feasibility of implementing a supervised exercise program, as well as to determine the barriers of recruiting and implementing a laboratory-based supervised exercise intervention for both AA and white CRC survivors. We hypothesized that exercise is beneficial for CRC survivors and that it may be more difficult to recruit AA CRC survivors.

METHODS

The study incorporated a mixed methods design. We completed two studies, one to determine the feasibility of recruiting CRC survivors into a 12-week supervised exercise program and a second incorporating semi-structured interviews to examine barriers to participating in an exercise intervention. One participant (white male) completed both the quantitative and qualitative studies. Both studies were approved by the University at Buffalo (UB) and the Roswell Park Comprehensive Cancer Center (RPCCC) Human Subjects Institutional Review Boards and written informed consent was obtained from all participants.

Quantitative Study

The quantitative component was carried out between October 2013 and December 2014.

Subjects

Recruitment of AA & white CRC survivors included mailings to patients from a comprehensive cancer center who were within 10 years of diagnosis, along with postings in the community and community-based oncology practices. The direct mailings consisted of identifying the institutions involved (UB and RPCCC) as well as the study goal, “identifying how exercise and overall fitness may benefit colorectal cancer survivors”. It was explained they would be given a free cardiopulmonary fitness evaluation, a personalized exercise program and they would be financially compensated for their time. Two weeks following the group mailings patients received a follow-up phone call from a research assistant to determine interest and to complete pre-screening. The flyers posted in the community included the same information. Participants in the exercise intervention were eligible for a modest compensation up to $200.

A sample size of 24 participants was proposed, including 14 AA and 10 whites, equally balanced by gender; n=5/group is sufficient to effectively determine a statistical difference in exercise capacity¹². Since this was a feasibility study sample size calculations were not applicable.

Inclusion criteria

Survivors of CRC who self-identified as AA or white, 45 to 74 years of age, any stage of disease, all active treatment completed (includes radiation, surgery and chemotherapy); anticipated life expectancy >12 months; able to walk unassisted ≥ 100 meters; no contraindications to aerobic exercise and no medical contraindication to participation.

Exclusion criteria

Self-reported lung disease, ischemic heart disease, congestive heart failure, and/or significant cardiac arrhythmias; inability to read/write/understand English language; limiting orthopedic, musculoskeletal, or psychological conditions (clinician discretion). Subjects were deemed ineligible for continued participation if baseline testing demonstrated: (1) Forced expiratory volume in one second (FEV₁) < 80% predicted, (2) FEV₁/forced vital capacity (FVC) ratio <0.70 or room air oxygen saturation levels <90%.

Design

Phase 2A single arm clinical trial of a 12-week supervised exercise intervention. The same investigators completed exercise testing at both baseline and at post-training (follow-up); exercise testing was conducted over two days separated by 3-4 days.

Exercise Intervention

Following completion of baseline assessment, participants reported to the laboratory twice per week, over a 12-week period to complete their supervised exercise sessions on a cycle ergometer (Ergoline Ergoselect100, Germany). For the first 4 weeks, exercise training began at intensities equal to 50-60% of baseline VO_2Peak for 20-25 minutes/session. For weeks 5-8, exercise intensity was increased to 60-70% of VO_2Peak and time was increased to 25-35 minutes. For weeks 9-12, exercise intensities were increased to 70-80% of baseline VO_2Peak and time was extended to 35-45 minutes. A 5-minute warm up and cool down were included for all sessions. Blood pressure and heart rate were recorded throughout.

The following laboratory tests were completed pre- and post-intervention

Peak and Sub-Maximal Exercise Test

A symptom limited progressive exercise test with electrocardiogram (ECG, Viasys, Yorba Linda, CA) was performed on an electronically braked cycle ergometer (Ergoline Ergoselect100, Germany) with a cardiopulmonary exercise testing system (Viasys, Vmax Encore 229, Yorba Linda, CA). Peak oxygen (average over last 30 sec) consumption was used to establish workloads for the 12-week cycling intervention program. The six-minute walk test (6MWT), was performed according to American Thoracic Society (ATS) guidelines¹³. The 6MWT is a valid and reliable measure (ICC r=0.93) that correlates with peak cardiopulmonary exercise capacity (VO_2peak, r=0.67) in cancer patients¹⁴.

Health-related quality of life was measured using the Short Form-36 (SF-36)¹⁵ and the Functional Assessment of Cancer Therapy Scale-Colorectal (FACT-C) measure¹⁶. The SF-36 is a widely used quality of life (QoL) measure and FACT-C is a validated measure assessing quality of life concerns among patient with colorectal cancer. Fatigue was measured using the Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-F) questionnaire¹⁷. The 18-item Brief Symptom Inventory (BSI) was used to assess self-reported symptoms as a measure of psychological distress in CRC survivors¹⁸. The Memorial Sloan Kettering Cancer Center Bowel Function Instrument (BFI), a validated 41-item questionnaire, was used to assess bowel function¹⁹.

Qualitative Study

The qualitative component was carried out between December 2014 and July 2015, in this case, primarily to further explain the less than optimal recruitment and retention in the intervention and discover specific issues with the protocol process. Using the eligibility criteria outlined above, CRC survivors were recruited from a local comprehensive cancer center. The same group of patients that received letters as stated above were re-contacted for the qualitative interviews; recruitment continued until we obtained saturation of responses in the thematic categories. Following oral consent, patients participated in one-on-one interviews, completed in a private office or on the phone, lasting 15-25 minutes. A semi-structured set of questions were asked regarding past and present attitudes and beliefs regarding exercise, willingness to participate in exercise/physical activity, and preferred types of activities. All interviews were captured using a digital audio recorder along with hand written notes. The audio-recorded interviews were transcribed and de-identified. Participants in the qualitative study were mailed a $10 gift card at the conclusion of the interview.

Qualitative findings were analyzed following methods developed and tested in prior research focused on contextualizing and tailoring responses to inform intervention development²⁰ using the PEN-3 Model. Structured interviews were transcribed verbatim and reviewed by two research team members using textual analysis with QSR NVivo version 10. We conducted triangulation to obtain consensus among investigators and resolve any inconsistencies through consensus and clarification of categories and themes. In addition, coded text was compared to demographic data, reported challenges from participants and published data on other studies of recruitment for minority and cancer survivors. The resulting categories were discussed with the remaining study team members. Responses were categorized into two of the PEN-3 domains (Relationships and Expectations and Cultural Empowerment) using a three by three table related to “Perceptions, Enablers or Nurturers” (Relationships and Expectations) and were further classified as “Positive, Existential or Negative” (Cultural Empowerment)²⁰. The third domain (Cultural Identity – Person, Extended Family, Neighborhood) was not included as the intervention was necessarily designed to be addressed for the cancer survivor-i.e., Person). For this PEN-3 analysis, Perceptions, Enablers and Nurturers (Relationships and Expectation domain) are defined as follows:

Perceptions = individually held knowledge, attitudes, values or beliefs stated by participants that are interpreted as facilitating (Positive) or hindering (Negative) personal motivation and decision-making related to exercise or participation in an exercise study²⁰.

Enablers = societal, systematic or structural influences that may enhance (Positive) or create barriers (Negative) to maintain or change the practice of exercise²⁰.

Nurturers= supportive (Positive) and/or discouraging (Negative) influences that a person may receive from significant others as it relates to exercise or study participation²⁰.

Following the categorization in this domain, responses were then determined to be Positive, Negative or Existential in relationship to the desired behavioral intervention. This PEN-3 Model of analysis incorporates dimensions of culture, health beliefs and behavior that can then be used to inform future interventions for behavior change^21,22. For this study, the goal of using these PEN-3 domains was to identify salient and potentially modifiable cultural and behavioral factors that were or were not being addressed by the intervention.

STATISTICAL ANALYSIS

Participant characteristics and outcomes are reported as means, medians and standard deviations for continuous variables; and as frequencies and relative frequencies for categorical variables. Levels of adherence with the exercise intervention and pre/post assessment and changes in SF-36, BSI-18, BFI, FACT-C, FACIT-F and exercise/physiologic parameters from baseline to the end of the 12-week intervention period were evaluated using paired t-tests or Wilcoxon signed rank tests if the data was not normally distributed. All analyses were conducted at a significance level of 0.05 and since this study was exploratory in nature, no adjustments were made for multiple hypothesis testing.

RESULTS

Intervention component

112 white patients from the RPCCC Cancer Registry were contacted (54 males, 58 female) via mailed letters; 25 (11 males, 24 female) patients were interested in participating and 5 patients consented to participate (4 males, 1 female). All five completed the 12-week program. In comparison, 126 AA patients from RPCCC were sent letters (54 males, 58 females); 15 (10 males, 5 females) patients were interested in participating and ten patients consented to participate (6 males, 4 females). Four males and two females never completed baseline testing (1-hernia repair, 4-medical/cardiovascular, 1-no response to calls), while four AAs (2 males, 2 females) started the training. The two females dropped out of the study shortly after starting (undisclosed medical reason and a non-related musculoskeletal injury). Finally, only two AA males completed the 12-week program.

The mean age ± SD of the 9 participants who initiated the intervention study were 56.9±6.7 years of age and were 3.6±2.3 years from diagnosis and 3.1±2.5 months from last treatment. They were 75% male and 77% were never smokers. Prior therapy included surgery (9/9, 100%) and chemotherapy (1/9, 11%). Baseline mean BMI 33.0±4.2 (kilograms/meter²). Among the seven (5 white/2 AA) participants who completed the 12-week cycling intervention (24 sessions) the overall session compliance rate was 96%.

Mean age ± SD for the five white (4 males and 1 female) and four AA (2 males and 2 females) participants averaged 55.8±58.3 and 58.3±8.5 years of age (p=0.62); participants were 2.0±1.6 vs.5.6±1.3 years from diagnoses (p=0.01) and 1.8±1.5 vs. 5.5±1.4 years from last treatment (p=0.01), whites and AAs, respectively. Baseline BMI was 32.4±4.6 and 33.8±4.1 for the five white and four AA (n=4) (p=0.63) participants. The two AA women who dropped out of the study were not demographically different from the seven participants who completed the intervention; however, they were the only smokers in the study. The AA participants had a 20% lower FEV₁ (p=0.01) as well as covered less distance in 6MWT at baseline (AAs, 484±38 feet versus whites 621±106 feet, p=0.045). Peak oxygen consumption averaged 22.5± 8.7 and 21.2±9.1 ml/kg/min for white and AA participants, respectively (p=0.68).

Peak and Sub-Maximal Exercise

Post-intervention assessment demonstrated a 17% increase in VO_2Peak p=0.011), a 23% increase in VCO₂ (p=0.007), a 32% increase in minute ventilation (p=0.010) and a 27% increase breathing frequency (p=0.020, table 1). Participants also achieved a higher workload (p=0.037) and a longer time to exhaustion (p=0.013) post-training. The distance covered during the 6MWT was unchanged (p=0.37). Body weight remained unchanged (pre- versus post-intervention).

Table 1.

Measures of Peak and Submaximal Exercise Pre- and Post-Intervention in Colorectal Cancer Survivors.

	Pre (n=7)	Post (n=7)	p-value
VO2peak (ml/kg/min)	21.7±7.5	24.8±6.7	0.011
VCO2peak (l/min)	2.1±0.5	2.7±0.8	0.007
Ve (l/min	64.0±14.7	84.9±25.3	0.010
Fb (breaths/min)	29.4±4.2	36.6±3.5	0.020
Workload (watts)	114±36	141±46	0.037
Time (min)	11.9±2.9	13.4±2.8	0.013
6MWT (m)	583±111	597±57	0.37

Values represent means ± SD pre and post-training. VO_2peak=peak oxygen consumption, VCO_2peak=peak carbon dioxide production, Ve = minute ventilation, Fb = breathing frequency, Time = time of peak exercise test, 6MWT=walking distance covered in 6 minutes. T-tests were used to test for differences. Two-tailed P-values <0.05 are considered significant.

Self-Report Questionnaires

As shown in table 2, the SF-36 physical component score significantly improved following the intervention (p=0.040), although the mental health score did not improve (p=0.444). No changes were observed in the self-reported scales assessing cancer-specific fatigue (FACIT-F, p=0.262), QoL assessed using the FACT-C (p=0.407), bowel function (BFI) subscales or total score (all >0.05) or the BSI (p=0.283).

Table 2.

Responses to Selected Survey Instruments Pre- and Post-Exercise in Colorectal Cancer Survivors

Questionnaires	Pre (n=7)	Post (n=7)	p-value
SF-36
Physical	48.3 ± 9.8	51.7 ± 7.2	0.040
Mental	53.4 ± 5.0	54.8 ± 3.9	0.444
FACT-C	108.9 ± 23.1	113.6 ± 15.7	0.407
FACIT-F	131.6 ± 25.8	136.7 ± 19	0.262
MSKCC Bowel Function
Frequency	22.9 ± 3.2	24.1 ± 3.7	0.438
Diet	15.6 ± 2.7	15.0 ± 5.0	0.618
Urgency	18.6 ± 1.7	16.4 ± 5.8	0.586
Total	72.9 ± 8.7	72.1 ± 11.6	0.938
BSI	2.9 ± 5.0	3.6 ± 5.6	0.283

Table 2 presents means ± SD from the QoL (SF-36), fatigue (FACIT-F) and symptom questionnaires (BFI, BSI, FACT-C) from the exercise study (n=7). The 36-Item Short Form Health Survey (SF-36) and subscales including physical and mental scores; Functional Assessment of Cancer Therapy-Colorectal Scale (FACT-C); Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-F); Memorial Sloan Kettering Bowel Function Instrument (BFI) and subscales including frequency, diet, urgency and the total score; and the Brief Symptom Inventory (BSI). Two-tailed p-values <0.05 were considered significant based on paired t-tests (FACT-C, FACIT, and BSI) or Wilcoxon signed rank tests (SF-36 and BFI) if the data was not normally distributed.

Formative findings from the Interviews

Table 3 summarizes demographic characteristics from the 30 participants in the qualitative study. One participant (white male) completed both the quantitative and qualitative studies. Based on ages and other characteristics, these participants appear to be representative of the CRC patient population at the cancer center and in the region and there were no differences by race.

Table 3.

Selected Demographic Characteristics of Participants in the Qualitative Component by Self-Reported Race in Colorectal Cancer Survivors (n=30)

Qualitative Demographics		African American (n=12)	White (n=18)	Total (n= 30)	p-value
Gender	Male	4 (33%)	7 (39%)	11 (37%)
	Female	8 (67%)	11 (61%)	19 (63%)	0.757
Age	40-49	0 (0%)	2 (11%)	2 (7%)
	50-64	7 (58%)	8 (45%)	15 (50%)
	65-74	3 (25%)	4 (22%)	7 (23%)
	75+	2 (17%)	4 (22%)	6 (20%)	0.627
Employment Status	Full time	1 (8%)	5 (28%)	6 (20%)
	Part time	2 (17%)	2 (11%)	4 (13%)
	Retired	5 (42%)	7 (39%)	12 (40%)
	Unemployed	3 (25%)	4 (22%)	7 (24%)
	Unknown	1 (8%)	0 (0%)	1 (3%)	0.547
Marital Status	Single	7 (59%)	3 (17%)	10 (33%)
	Married	3 (25%)	10 (56%)	13 (44%)
	Divorced	0 (0%)	1 (5%)	1 (3%)
	Widowed	1 (8%)	4 (22%)	5 (17%)
	Unknown	1 (8%)	0 (0%)	1 (3%)	0.081
Income	<$25,000	2 (17%)	4 (22%)	6 (20%)
	$25,000-34,999	0 (0%)	1 (6%)	1 (3%)
	$35,000-49,999	2 (17%)	1 (6%)	3 (10%)
	$50,000-74,999	2 (17%)	7 (38%)	9 (30%)
	$75,000-99,999	0 (0%)	0 (0%)	0 (0%)
	$100,000-149,000	0 (0%)	4 (22%)	4 (14%)
	$150,000+	1 (7%)	0 (0%)	1 (3%)
	Unknown	5 (42%)	1 (6%)	6 (20%)	0.069
Highest Level of Education	Some High School	1 (8%)	2 (11%)	3 (10%)
	High School/GED	4 (34%)	5 (28%)	9 (30%)
	Some College	4 (34%)	0 (0%)	4 (13%)
	Associates Degree	0 (0%)	4 (22%)	4 (13%)
	Bachelor’s Degree	1 (8%)	5 (28%)	6 (21%)
	Graduate DegreeUnknown	1 (8%)1 (8%)	2 (11%)0 (0%)	3 (10%)1 (3%)	0.066

p-values <0.05 were considered significant.

Table 4 shows the distribution of the number of responses categorized as Positive and Negative Perceptions, Enablers and Nurturers, as they apply to exercise and participation in an organized exercise study for CRC survivors. As shown in Table 4, only one response was categorized as Existential, which refers to values and beliefs that are practiced within a specific culture but pose no threat to exercise or study participation²⁰. Thirty-seven percent more of the participant responses were categorized within the Positive column (n=174) compared to barriers or negative influences on exercise (n= 127), so this offers many specific attitudes, perceptions and influences on which to focus to promote exercise among potential CRC survivors.

Table 4.

PEN 3-Examples of Positive, Existential and Negative Responses Expressed by Colorectal Cancer Survivors by Category

	POSITIVE	EXISTENTIAL	NEGATIVE
PERCEPTIONS	My weight really has to be controlled with my health and stuff, so I can’t really gain much or lose too much. Improved physical mobility and health and strength. That kind of thing? It has helped me out. I’m not using my cane as much because I’ve had balance issues. I have MS, so it’s a combination of a lot of things, so I been trying to strengthen myself in terms of that. I walk. I golf. I take care of things; the lawn whatever. I just started…this is no big thing, but I just started doing some yoga things. I want the muscle tone back and everything. Building up your actual physical strength and also your confidence. It makes you feel a lot stronger in dealing with life. Well, I’d like to be able to do more things because I think not being active, my hip is getting worse because even going up and down stairs is tough. I’d like to be able to do more so that my hip would move better. Probably heart things and maybe things with arthritis because I think when you’re exercising and stuff like that, I think that would be good. I enjoy walking.		It’s just not me. I’m more of a, well not a loner, but I’m kind of to myself. Laziness. That’s just about it. Physically I’m fine. I had been exercising when I was diagnosed and then after that I kind of stopped with all the treatment. Then my knees just got so bad I couldn’t. I have two hips that are bone on bone and a bad back, so it’s low. I have to get them fixed. I’m disabled. I can’t work. I don’t have any barriers other than being lazy. It’s just the starting and making it a habit. I understand why you’re doing this, but I don’t think I would fit in with a regular group. I wear a full leg brace. How would I fit in? If it’s free and it’s easily accessible by bus or by me walking. I don’t want to go all over Buffalo. Because I don’t really don’t like doing exercise. Not really. I like to read and I like to play my games on the Kindle.
ENABLERS	In the summertime I go to a lake and I swim out far Well, I’m walking, shoveling snow, doing the lawn or cleaning around the house. Cleaning up the kitchen, wiping down my dining room table, making up the beds and things of that nature. Maybe golf come spring or summer. Swimming when the pools open or just playing volleyball or something. I like to walk. I walk everywhere in the snow. If it’s free and it’s easily accessible by bus or by me walking. I cut my grass every weekend. So, that’s very strenuous. I do all my cleaning and cooking. I have arthritis, so that right there really motivates me to exercise more and move around more and then my joints won’t get stiff. Well, I like walking the stairs. Everything in the summer; swimming, softball.	Well, not necessarily, but it depends on the location.	Moderate. I’m just waiting for the snow to melt. I’m kind of stuck in the house I don’t drive; that’s the biggest. I sometimes watch my grandson, so my time is…I don’t know sometimes until the last minute. It’s hard to make schedules that way. Those are probably the top two. I don’t think I have any other. Well, because of winter, moderate. I can’t get out much. Depends on where it’s at. If it’s a long drive then I couldn’t do it. Well, the way the weathers been, I haven’t been. I craft. I scrapbook and that sort of thing. I belong to a Scrabble group that meets once a month and I belong to a book club. All not active-type activities. pain. I haven’t had it. I’m not saying that I don’t expect it, but when my knees were bad, I just couldn’t do it. Because of the time restraints. I wouldn’t like to take that time out to go do anything like that. I would have to get someone to take me.
NURTURERS	I grocery shop. I cook. I clean the house. I babysit, so I chase after the little one. Me and my wife together we walk the whole neighborhood. We didn’t used to walk. We just started. …my granddaughter before I got sick I would take her to the playground and swing her on the swings, but now I can’t even pick her up. I’m scared to do a whole lot of stuff because I don’t know what happens when a hernia busts. Yes. I’ve got a six year-old and a two and a half year-old. Participating with other people who are trying to do the same thing that I’m doing that I would like to do. I like walking a lot. I walk my neighborhood with my friends,		family commitments I have a mother who I care for. I’m her primary caregiver. He said, “Usually with xrays like this, we would see you in a walker or a wheelchair.” Only because of my knees; I got titanium in both knees and my orthopedic surgeon wants to do my hips. He basically said, “I’m not sure how you’re walking.”

Responses from the structured interviews are summarized and quantified by theme/topic and demographics in Table 5 to illustrate the physical activities of interest, as well as perceived barriers and benefits of physical activity. Overall, participants uniformly offered a strong endorsement for cardiovascular training based on the number of comments expressing an interest in aerobic exercise (e.g. aerobic vs. resistance exercise) (46/78 total responses). Current types of physical activities reported included housework (13/28 responses reported by females and whites) and yardwork (13/28 responses reported by males and whites). Specific barriers to physical activity included illness and injury (28/67 total responses), weather/location (14/67 responses) and lack of motivation (14/67 responses). The top benefits to exercise included improved physical health (27/42 responses) and weight control (16/42 responses). Topics of interest for potential education classes included nutrition & diet (17/28 responses), exercise (6/28 responses) and cancer prevention (5/28 responses). Females and younger participants provided the largest number of responses, while little to no differences were noted between sex, race or by age category.

Table 5.

Structured Interview Qualitative Themes expressed by Colorectal Cancer Survivors by Gender, Race and Age Category

	Number of References
	Overall	Gender		Race		Age
Name		Female	Male	AA	White	40-64 years	65+ years
1. Physical activities of interest	78	49	29	36	42	53	25
a. Cardio	46	28	18	18	28	29	17
b. Strength Training	9	4	5	4	5	7	2
c. Hobbies	22	16	6	10	12	14	8
d. Yoga	6	5	1	0	6	5	1
e. Other	8	2	6	6	2	5	3
f. None	0	0	0	0	0	0	0
2. Daily Activity	28	15	13	7	21	10	18
a. Housework	13	9	4	3	10	5	8
b. Baking and Cooking	6	6	0	1	5	3	3
c. Laundry	1	1	0	1	0	0	1
d. Yardwork and Outdoor	13	3	10	1	12	4	9
e. Errands and Shopping	2	1	1	1	1	1	1
f. Babysitting	1	1	0	0	1	1	0
3. Barriers to physical activity	67	48	19	19	48	37	30
a. Lack of motivation	14	8	6	0	14	4	10
b. Transportation	7	7	0	5	2	3	4
c. Time	8	4	4	2	6	7	1
d. Illness and Injury		21	7	7	21	16	12
e. Tired	0	0	0	0	0	0	0
f. Weather	8	5	3	4	4	6	2
g. Location	6	5	1	2	4	2	4
h. Motivation	0	0	0	0	0	0	0
i. Money	2	2	0	1	1	2	0
j. Nothing	1	0	1	1	0	1	0
4. Benefits to exercise	42	28	14	17	25	24	18
a. Health	27	18	9	10	17	19	8
b. Weight	16	10	6	6	10	9	7
c. Diet	2	1	1	0	2	1	1
d. Reduce Fatigue	2	1	1	1	1	1	1
e. Cardiovascular Health	4	2	2	2	2	1	3
f. Unsure	1	1	0	1	0	0	1
5. Nutrition	12	7	5	6	6	8	4
6. Health Topics of Interest	28	20	8	10	18	17	11
a. Nutrition and Diet	17	13	4	6	11	11	6
b. Health	4	4	0	1	3	1	3
c. Cancer Prevention	5	2	3	2	3	3	2
d. Different cancer types	2	2	0	0	2	2	0
e. Exercise	6	5	1	3	3	6	0
f. Aging	1	1	0	0	1	1	0
g. None	1	0	1	1	0	0	1
7. Reasons to not participate in research	18	12	6	5	13	8	10
a. Like to exercise alone	1	1	0	0	1	1	0
b. Location	5	2	3	0	5	3	2
c. Time constraints	3	2	1	2	1	1	2
d. Not interested	5	3	2	1	4	1	4
e. Transportation	3	3	0	1	2	2	1
f. Health issues	4	4	0	2	2	2	2

AA, African American. Note: Numbers in columns signify the number of references/quotes under each category, not sample size; bolded numbers designate category subtotals.

As illustrated in Table 4, many Positive Perception responses reflected the recognition by survivors that exercise provided important benefits for their health, and more activity was a goal. Positive Enabler examples included several issues that are seasonal for Western New York, and prove to be important in any type of exercise study, especially with regard to home-based activities. Several people also commented about equipment they had that could be included in a planned exercise program. Negative Perceptions that will need to be addressed in future intervention development, and may have negatively influenced recruitment include co-morbidities/physical limitations, expressions of “laziness” and perceptions that it is too much trouble to get to a location to exercise:

“Laziness. That’s just about it. Physically I’m fine.” (Female, White)

“I had been exercising when I was diagnosed and then after that I kind of stopped with all the treatment. Then my knees just got so bad I couldn’t. For that four or five years I was in treatment, nothing was going on.” (Female, White)

“Because of time restraints. I wouldn’t like to take that time out to go do anything like that.” (Male, African American)

Likewise, Negative Enablers (structural/systematic factors) such as weather and transportation issues outnumbered Positive Enablers (60 vs. 49), and are challenges that will need to be addressed (Table 4).

DISCUSSION

Our primary objective was to explore the feasibility of recruiting AA and white CRC survivors into a laboratory-based supervised exercise program, and secondarily to utilize qualitative methods to explore factors that hinder or promote participation in structured exercise. The results from this study confirm that recruiting CRC survivors into a laboratory-based supervised exercise program using direct mail and flyers was challenging, regardless of race, yet those that completed the intervention (both AAs and whites) demonstrated measurable improvements in cardiorespiratory fitness and QoL.

This study demonstrated that a 12-week cycling intervention improved peak but not submaximal exercise performance. The improvement in peak exercise (Δ = 3.2±2.3 ml/kg/min) is consistent with other CRC studies⁸, including those incorporating a 12-week combined resistance and aerobic training component²³. Overall, these improvements are indicative of improved cardiorespiratory health (VO_2Peak), a variable known to predict mortality and survival in CRC²⁴. Besides low power, training specificity and that, the 6MWT is a self-paced test; it is unknown why the 6MWT did not improve post-training.

A meta-analysis has recently shown that exercise is one of the best interventions to reverse fatigue in cancer survivors²⁵. Despite exercise being more effective than pharmacological interventions, the mechanisms describing the fatigue-reducing benefit of exercise training, and what type of training is optimal to attenuate fatigue is unknown. The inability to improve fatigue following the 12-week cycling intervention is similar to previous reports in home-based exercise studies in CRC survivors^8,26 and may represent the smaller sample size or less fatigue initially, since improvements in fatigue may be associated with baseline fatigue and fitness level²⁶.

Observed improvements in the physical health score of the SF-36 following the supervised exercise intervention represents an important finding because some²³, but not all CRC exercise interventions were able to improve QoL post-intervention^8,26. This lack of improvement in the QoL measure was attributed to limited statistical power in one study²⁷, whereas, Courneya et. al., demonstrated improvements in those who were more active, but not by group randomization²⁶. Because QoL is strongly associated with physical activity post-diagnoses²⁶ this may explain why the improvements seen in the current study are closely aligned with studies incorporating laboratory based/supervised exercise programs^23,27 compared to unsupervised home-based programs^8,26 where patients are not as active. Also, CRC survivors may have multiple comorbidities limiting their ability to perform exercise²⁸, which represents a potential a constraint to engaging survivors in exercise, especially unsupervised programs.

Recruitment for this exercise intervention was less than optimal. Letters were mailed out to 238 (126 AA and 112 white) CRC survivors, 9 (3.8%) agreed to participate and only 7 (2.9%) completed the 12-week intervention; no differences were observed by race. Challenges in recruiting survivors (AA and white) appears to be common in many, if not all, CRC exercise studies. Studies using a similar recruitment approach (letters) also had a comparable enrollment rate (9-10%)^27,23, in comparison to other studies that were somewhat more successful, accruing (~30-40%) through recruitment which included a combination of the internet/social media, letters from oncologists, personal contact, radio and printed advertisements^8,26. It is possible we did not employ enough strategies to engage survivors, however, a recent review demonstrated letters to participants successfully generated AA accruals onto studies (range=10-80%)²⁹. Rewarding participants with fitness evaluations and personalized exercise programs as well as financial incentives do not appear to be important variables when recruiting AA and white CRC survivors. Although, each study and location represents unique challenges, it appears that a combination of community outreach, along with a multi-center approach may help to maximize recruitment. Despite low accrual, adherence with scheduled sessions (>95%) was similar to that observed in other CRC studies of exercise intervention^23,27.

Results from the qualitative component of our research suggest that multiple challenges and perceptions among survivors also need to be addressed to increase participation rates into exercise interventions. In hindsight, incorporating more qualitative research initially would have perhaps improved the approach. Findings from the current study suggest that future interventions addressing patient co-morbidities, along with convenient times and locations for scheduled training may be more likely to engage participants. These barriers are consistent with those previously reported by white CRC survivors who noted a lack of time, treatment-related side effects and fatigue as three of the most common barriers to exercise⁶. “Lack of motivation” to exercise was stated as a barrier by many participants in the current study which may be interpreted as potentially related to fatigue and/or medical co-morbidities or even dysphoria/depression. The lack of motivation is consistent with the published literature which states <30% of CRC survivors meet activity guidelines post-treatment^6,30. However, a majority of CRC survivors are physically incapable of meeting exercise guidelines immediately following curative care (<6 months)²⁸. Therefore, factors such as time from last treatment, medical co-morbidities and level of motivation should be considered when attempting to increase physical activity in cancer survivors. Bowen et al. (2009) have suggested a framework for feasibility studies which includes acceptability, demand, implementation, practicality, adaptation, integration, expansion and limited-efficacy testing. Our results indicate shortcomings related to several of these domains which warrant consideration when similar studies are attempted.

Future steps will incorporate these domains, the qualitative findings, and resulting discussion and suggestions from patients. For example, perceptual challenges about exercise may be better addressed through direct physician referrals from clinical oncology programs more immediately after treatment completion. Unfortunately, few CRC survivors reported recalling conversations with health care providers regarding physical activity; those that were able to recall were more likely to report greater levels of activity post-treatment³¹. It is also notable that all CRC patients from the current qualitative study recognized the need for and benefits associated with exercise and are generally willing to participate if they are motivated and supported by others. Efforts to improve activity in CRC survivors are important due to the fact many share common risk factors with other obesity-related diseases such as diabetes, heart disease and cancers. Thus, a personalized approach to exercise and self-management in the patient’s survivorship care plan may help with lifestyle modifications regardless of race³².

The strengths of this manuscript include the mixed methods design, which permitted a more comprehensive assessment of patient preferences and opinions. The primary limitations include the limited number of participants, making it difficult to develop conclusions, however we did note statistically and clinically relevant improvements following the intervention component. Our findings suggest that challenges recruiting are multifactorial and include the study timeline, limited number of AA patients at many facilities (including our own), variable levels of patient interest and a reliance upon a single clinical center.

In summary, this feasibility study confirms that participation in a supervised exercise program improves cardiorespiratory performance and QoL, but not fatigue and body weight in CRC survivors. Regardless of race, it is challenging to use direct mailings and flyers when accruing CRC survivors to a supervised exercise program, despite the success of this method in healthy populations of the same age and race. This study also adds qualitative insights into the QoL of CRC survivors and their relationships between the clinicians and scientists who are attempting to reduce their mortality risk with exercise interventions.