A randomized controlled trial of a multiple health behavior change intervention delivered to colorectal cancer survivors: Effects on sedentary behavior

Authors

  • Brigid M. Lynch PhD,

    Corresponding author
    1. Physical Activity Laboratory, Baker IDI Heart and Diabetes Institute, Melbourne, Victoria, Australia
    2. Melbourne School of Population and Global Health, The University of Melbourne, Melbourne, Victoria, Australia
    • Corresponding author: Brigid M. Lynch, MD, Physical Activity Laboratory, Baker IDI Heart and Diabetes Institute, PO Box 6492, St Kilda Rd Central, Melbourne VIC 8008 Australia; Fax: (011) 61 3 8532 1100; brigid.lynch@bakeridi.edu.au

    Search for more papers by this author
  • Kerry S. Courneya PhD,

    1. Department of Physical Education and Recreation, University of Alberta, Edmonton, Alberta, Canada
    Search for more papers by this author
  • Parneet Sethi BSc,

    1. Physical Activity Laboratory, Baker IDI Heart and Diabetes Institute, Melbourne, Victoria, Australia
    Search for more papers by this author
  • Tania A. Patrao MIPH,

    1. Menzies School of Health Research, Brisbane, Queensland, Australia
    Search for more papers by this author
  • Anna L. Hawkes PhD

    1. School of Public Health and Social Work, Queensland University of Technology, Brisbane, Queensland, Australia
    2. School of Public Health, Tropical Medicine and Rehabilitation Sciences, James Cook University, Townsville, Queensland, Australia
    Search for more papers by this author

Abstract

BACKGROUND

Sedentary behavior may independently contribute to morbidity and mortality among survivors of colorectal cancer. In the current study, the authors assessed whether a telephone-delivered multiple health behavior change intervention had an effect on the sedentary behavior of recently diagnosed colorectal cancer survivors.

METHODS

A total of 410 participants were recruited through the Queensland Cancer Registry and randomized to the health coaching (intervention) or usual-care (control) group. Eleven health coaching sessions addressing multiple health behaviors, including sedentary behavior, were delivered over a period of 6 months. Data were collected at baseline (before randomization), at 6 months, and at 12 months via a telephone interview.

RESULTS

At 12 months, there was a significant decrease noted in the hours per day of sedentary time in both the health coaching (−1.21; 95% confidence interval [95% CI], −1.71 to −0.70) and usual-care groups (−0.55; 95% CI, −1.06 to −0.05), but the between-group difference was not found to be statistically significant (−0.65; 95% CI, −1.37 to 0.06 [P = .07]). In stratified subgroup analyses, the multiple health behavior change intervention was found to have a significant effect on total sedentary time (hours/day) at 12 months in survivors of colorectal cancer who were aged > 60 years (−0.90; 95% CI, −1.80 to −0.01 [P = .05]), male (−1.33; 95% CI, −2.44 to −0.21 [P = .02]), and nonobese (−1.10; 95% CI, −1.96 to −0.25; [P = .01]).

CONCLUSIONS

Incorporating simple messages about limiting sedentary behaviors into a multiple health behavior change intervention was found to have modest effects on sedentary behavior. A sedentary behavior-specific intervention strategy may be required to achieve substantial changes in sedentary behavior among colorectal cancer survivors. Cancer 2014;120:2665–2672. © 2014 American Cancer Society.

INTRODUCTION

Colorectal cancer is associated with significant decrements in health status and an increased risk of death from noncancer causes.[1] However, modifiable health behaviors can redress such morbidity. For example, regular participation in moderate- to vigorous-intensity physical activity is associated with prolonged survival,[2, 3] diminished treatment side effects, and an enhanced quality of life.[4, 5] Sedentary behavior (commonly conceptualized as “sitting time”) is a health risk that is additional to, and distinct from, the hazards of too little moderate- to vigorous-intensity physical activity.[6] Sedentary behavior is associated with adiposity, insulin resistance, and markers of inflammation, and therefore could plausibly contribute to adverse cancer outcomes and to the development of comorbid chronic disease.[7]

To the best of our knowledge, an Australian prospective study of 1966 survivors of colorectal cancer provided the first evidence that sedentary behavior has deleterious associations with health outcomes in survivors of cancer.[8-10] Watching ≥ 5 hours versus ≤ 2 hours of television per day was associated with a mean increase in body mass index (BMI) of 0.71 kg/m2 over approximately 18 months,[10] a 16% lower total quality-of-life score,[9] and de novo ischemic heart disease.[8] Sedentary behavior has also been linked with colorectal adenoma recurrence in men in a US study using a pooled sample of participants from two randomized, double-blind, placebo-controlled phase III clinical trials (The Wheat Bran Fiber Study and the Ursodeoxycholic Acid Trial).[11]

To the best of our knowledge, only 1 study to date has considered the association between sedentary behavior and colorectal cancer survival.[12] In a recently published analysis, 2293 men and women with nonmetastatic colorectal cancer were identified within the Cancer Prevention Study II Nutrition Cohort.[12] Self-reported leisure-time sitting was ascertained at baseline (mean, 7 years before diagnosis) and at routine follow-up time points. The first assessment conducted after the diagnosis of colorectal cancer was used as the postdiagnosis measure of sitting time (mean, 2 years after diagnosis). Prediagnosis sitting time was associated with all-cause mortality (relative risk [RR] for < 3 hours/day vs ≥ 6 hours/day, 1.36; 95% confidence interval [95% CI], 1.10-1.68) whereas postdiagnosis sitting time was associated with all-cause (RR, 1.27; 95% CI, 0.99-1.64) and colorectal cancer-specific (RR, 1.62; 95% CI, 1.07-2.44) mortality.

Hence, sedentary behavior appears to be a modifiable factor that independently contributes to morbidity and mortality among survivors of colorectal cancer. Determining methods of successfully decreasing sedentary behavior, in addition to increasing physical activity, is an important step for colorectal cancer control and improved health outcomes. In the current study, we assessed whether a telephone-delivered multiple health behavior change intervention that focused primarily on physical activity, diet quality, weight management, and alcohol and tobacco use but included a component addressing sedentary behavior had an effect on sedentary behavior in a sample of recently diagnosed colorectal cancer survivors. We also examined intervention effects in selected subgroups based on clinically relevant indicators (age, sex, site, obesity, and physical activity status).

MATERIALS AND METHODS

Study Design

The methods of the CanChange trial have been described previously.[13] The primary outcomes of the trial included physical activity, health-related quality of life, and cancer-related fatigue. The intervention increased physical activity levels significantly (compared with those receiving usual care, patients in the intervention group were more likely to meet physical activity guidelines of ≥ 150 minutes/week at 12 months [16.4% vs 9.2%; P = .047]), but appeared to have limited effects on health-related quality of life or cancer-related fatigue.[13]

In brief, this study was a 2-armed randomized controlled trial evaluating a 6-month telephone-delivered multiple health behavior change intervention commencing within 12 months of a primary colorectal cancer diagnosis (n = 205 patients) compared with usual care (n = 205 patients). To be eligible for this trial, potential participants could not have any metastatic disease, had no medical conditions that would limit adherence to an unsupervised physical activity program (as confirmed by their referring physician), and did not meet ≥ 1 of the following health recommendations: 150 minutes of moderate- to vigorous-intensity physical activity per week14; 2 servings of fruit per day and 5 serving of vegetables per day15; or a body mass index < 25 kg/m.[2, 16] Ethics approval was obtained from The University of Queensland's Behavioural and Social Sciences Ethical Review Committee, and written informed consent was obtained for all procedures.

Study Conditions

Health coaching (intervention)

The intervention, which was based on acceptance commitment therapy,[17] has previously been described in detail.[18] Acceptance commitment therapy is an empirically based third-generation cognitive behavioral approach that uses acceptance and mindfulness strategies as well as commitment and behavior change strategies to produce psychological flexibility. It uses specific strategies to overcome internal barriers to making lifestyle improvements by emphasizing the role of emotions and thoughts in the self-regulation of health behaviors.[17, 19]

The intervention included: 1) 11 telephone-delivered health coaching sessions over a 6-month period delivered by nurses, psychologists, or health promotion practitioners (“health coaches”); 2) a participant handbook; 3) regular motivational postcard prompts; 4) a pedometer; and 5) the study quarterly newsletter. The average duration of the telephone sessions was 31 minutes, and 81% of participants randomized to health coaching received at least 6 of 11 telephone sessions. The telephone sessions addressed the cancer experience, colorectal cancer-related symptoms, and strategies to improve health behaviors consistent with Australian government recommendations[14-16, 20, 21] and individual goals. Although reducing sedentary behavior was not the primary focus of the intervention, the participant handbook emphasized “limiting sedentary habits such as watching television,” and made practical suggestions for replacing sedentary behaviors with light-intensity activities such as “walking instead of driving,” “getting off the bus 1 stop early and walking the extra distance,” and “walking around while talking on the cordless or mobile telephone.” These recommendations were discussed and reinforced during each of the telephone-delivered health coaching sessions.

Usual care (control)

At baseline, usual-care participants received freely available educational brochures produced by Cancer Council Australia, including 1 brochure regarding understanding colorectal cancer and 3 brochures concerning cutting cancer risk, diet, and physical activity, but no information relating to reducing sedentary behavior. Participants also received a quarterly study newsletter to enhance participant retention, and were contacted for all follow-up assessments.

Data Collection

Data were primarily collected through a computer-assisted telephone interview by staff who were blinded to group assignment. Interviews were conducted at baseline (before randomization) and at 6 months and 12 months (study endpoint).

Sedentary Behavior Assessment

Estimates of time spent in sedentary behaviors were reported during computer-assisted telephone interviews at each time point. Participants were asked to recall how many days they engaged in the sedentary behavior over the past 7 days and how many hours and minutes they spent in the sedentary behavior. The sedentary behaviors assessed were driving or riding in a car, television or video watching, computer use for leisure, video games, reading, talking on the telephone, and sitting to talk with friends (not on the telephone) or listen to music. These items were adapted from a sedentary behavior measure previously validated in a community sample.[22] Two summary scores were considered: screen time (the sum of reported time spent playing video games, watching television, and using a computer for leisure) and total sedentary time (the sum of reported time spent driving or riding in a car; total screen time; and time spent sitting and reading, talking on telephone, and talking with friends or listening to music).

Stratified Subgroup Analyses

We examined the intervention effects within subgroups of 5 variables, selected on the basis of their clinical relevance to physical function among survivors of colorectal cancer,[23-25] and findings from studies in nonclinical populations.[26, 27] Age (≤ 60 years vs > 60 years), sex, obesity (BMI < 30 kg/m2 vs ≥ 30 kg/m2), and physical activity status (≥ 150 minutes/week vs < 150 minutes/week) were assessed during the telephone interview at baseline; information regarding cancer site (colon vs rectum) was collected from cancer registry records at the time of recruitment.

Statistical Analyses

The sample size (N = 410) provided 80% power (5% significance, 2-tailed; 20% attrition) to detect a moderate standardized effect size d of 0.35 on the primary outcomes.[18] The trial was not originally powered to detect interactions or subgroup effects. Time spent in sedentary behaviors was summarized across the health coaching and usual-care groups. The Student t test was used to check for statistically significant differences between groups, with significance set at P < .05.

Intervention effects at 6 months and 12 months were assessed using linear mixed models for continuous variables. This model included random effects associated with the unit of analysis (participants). The model also allows the residuals associated with the longitudinal measures on the same unit of analysis to be correlated. All available data were used, and all participants were analyzed according to their random allocation to the groups. Models were adjusted for age because patients in the health coaching group were significantly younger than those in the usual-care group (P < .05).[13]

Intervention effects were examined across television viewing time and the 2 summary scores (screen time and total sedentary time). Reported time spent in the other sedentary behaviors was too low to expect any intervention effect. Intervention effects for total sedentary time were also examined across age, sex, obesity, physical activity, and cancer site categories. Results are presented as the mean difference in hours per day (95% CI). All analyses were conducted using Stata statistical software (version 12; StataCorp, College Station, Tex).

RESULTS

The flow of participants through the current study has been reported previously.[13] In brief, an initial sample of 1410 survivors of colorectal cancer was identified through the Queensland Cancer Registry and assessed for eligibility. From this group, 410 individuals (29%) were randomly assigned at a ratio of 1:1 to the health coaching or usual-care group. At 12 months, retention was 78% for health coaching and 80% for usual-care participants (P = .47).

Just less than one-half of the study participants were female (46%), 74% were aged > 60 years, 42% were overweight, and 20% were obese. The majority of participants (77%) were married or in a de facto relationship and 67% were in paid employment. In terms of disease-specific characteristics, 76% of participants were not undergoing any treatment, 12% had a permanent or temporary stoma, and 66% of the participants had colon cancer. Participants were characterized by a high prevalence of poor health behaviors. Only 12% were classed as sufficiently active (≥ 150 minutes/week of leisure-time physical activity) and 64% were overweight (BMI of ≥ 25.0 kg/m2). A significant percentage of participants consumed less than the recommended intake of fruit (39%) or vegetables (76%) per day, 23% consumed > 2 standard drinks per week, and 5% were current smokers. Although participants were randomly allocated to conditions, there was a statistically significant difference in age observed between the intervention and usual-care arms of the trial (65 years vs 68 years; P = .005). Therefore, the models were adjusted for age to account for this imbalance.

Table 1 presents the comparison of the health coaching and usual-care groups with regard to baseline sedentary behavior. No significant differences between the health coaching and usual-care participants were found.

Table 1. Sedentary Behaviors (Hours/Day) at Baseline by Health Coaching and Usual-Care Groups (n=410)a
 Health Coaching (n=205)Usual Care (n=205)P for Difference
  1. a

    Data are shown as the means (standard deviations).

  2. b

    Screen time includes time spent playing video games, watching television or videos, and using the computer.

  3. c

    Total sedentary time includes driving or riding in a car, total screen time, time spent sitting and reading, talking on the telephone, talking with friends, or listening to music.

Driving or riding in a car0.59 (1.63)0.42 (0.64).15
Watching television or videos2.73 (1.81)2.63 (1.62).56
Using a computer for leisure0.49 (1.12)0.38 (0.76).24
Playing video games0.01 (0.04)0 (0.02).44
Reading1.00 (1.28)1.20 (1.25).12
Talking on the telephone0.16 (0.37)0.16 (0.38).94
Talking or listening to music1.17 (1.21)1.22 (1.04).69
Screen timeb3.22 (2.11)3.00 (1.78).27
Total sedentary timec6.15 (3.45)6.01 (2.72).63

Main Effect of the Intervention on Sedentary Behavior

The main effects of the multiple health behavior change intervention on sedentary behavior in the current sample of colorectal cancer survivors are presented in Table 2. From baseline, there was a significant decrease in the number of total hours per day spent in sedentary time in the intervention group at 6 months (−0.65; 95% CI, −1.14 to −0.15 [P = .01]) and 12 months (−1.21; 95% CI, −1.71 to −0.70 [P < 0.01]) and in the usual-care group at 12 months (−0.55; 95% CI, −1.06 to −0.05 [P = .03]). There were no between-group differences noted with regard to change in total sedentary time (adjusted for age) at 6 months (−0.21; 95% CI, −0.91 to 0.49 [P = .56]) or 12 months (−0.65; 95% CI, −1.37 to 0.06 [P = .07]).

Table 2. Effects of a Multiple Health Behavior Change Intervention on Sedentary Behavior in Survivors of Colorectal Cancer in Queensland, Australiaa
 Within-Group Change (95% CI)Adjusted Between-Group Change (95% CI)b
 0 to 6 Months0 to 12 Months0 to 6 Months Change0 to 12 Months Change
  1. Abbreviation: 95% CI, 95% confidence interval.

  2. a

    The mean estimated absolute change (95% CI) and the mean estimated net difference for the change (95% CI) at 6 and 12 months for the health coaching and usual-care groups (health coaching group at 6 months [n = 171] and 12 months [n = 159] and usual-care group at 6 months [n = 176] and 12 months [n = 163]).

  3. b

    Between-group change represents the within-group change in the health coaching group minus the within-group change in the usual-care group, adjusted for age.

  4. c

    P < .05.

Watching television/videos, (h/d)    
Usual-care group−0.22 (−0.43 to 0.00)−0.29 (−0.52 to −0.07)c  
Intervention group−0.36 (−0.59 to −0.15)c−0.49 (−0.71 to −0.26)c−0.15 (−0.46 to 0.16)−0.19 (−0.51 to 0.12)
Screen time (h/d)    
Usual-care group−0.17 (−0.43 to 0.07)−0.24 (−0.50 to 0.02)  
Intervention group−0.41 (−0.67 to −0.16)c−0.56 (−0.83 to −0.30)c−0.23 (−0.60 to 0.13)−0.33 (−0.70 to 0.05)
Total sedentary time (h/d)    
Usual-care group−0.44 (−0.93 to 0.05)−0.55 (−1.06 to −0.05)c  
Intervention group−0.65 (−1.14 to −0.15)c−1.21 (−1.71 to −0.70)c−0.21 (−0.91 to 0.49)−0.65 (−1.37 to 0.06)

For screen time and television viewing time, no changes were detected in the usual-care group at either 6 months or 12 months, whereas decreases were observed in the intervention group at both time points (Table 2). There were no between-group differences noted for either screen time or television viewing time at either 6 months (screen time, −0.23 [95% CI, −0.60 to 0.13; P = .20] and television viewing time, −0.15 [95% CI, −0.46 to 0.16; P = .34]) or 12 months (screen time, −0.33 [95% CI, −0.70 to 0.05; P = .09] and television viewing time, −0.19 [95% CI, −0.51 to 0.12; P = .23]).

Stratified Subgroup Analyses of Intervention on Total Sedentary Time

We examined the interaction effect on change in total sedentary time, stratified across categories of 5 clinically relevant variables (Table 3). Significant effects were noted within the subgroups of age, sex, and obesity, but there was no significant intervention effect observed in the subgroups of cancer site or physical activity. A significant between-group change in total sedentary time at 12 months was present among participants aged > 60 years (−0.90; 95% CI, −1.80 to −0.01 [P = .05]), but there was no effect noted among younger survivors of colorectal cancer. Similarly, the intervention reduced total sedentary time at 12 months for men (−1.33; 95% CI, −2.44 to −0.21 [P = .02]) but not for women. There was a significant effect at 12 months among nonobese participants (−1.10; 95% CI, −1.96 to −0.25 [P = .01]). No between-group effects were observed at the intermediate endpoint (6 months) in any subgroup analyses, except for among the sufficiently active participants (−1.71, 95% CI, −3.29 to −0.13 [P = .03]).

Table 3. Stratified Subgroup Analyses of the Effects of a Multiple Health Behavior Change Intervention on Total Sedentary Time (Hours/Day) in Survivors of Colorectal Cancer in Queensland, Australia
 Within-Group Change (95% CI)Adjusted Between-Group Change (95% CI)
 0 to 6 Months0 to 12 Months0 to 6 Months Change0 to 12 Months Change
  1. Abbreviations: 95% CI, 95% confidence interval; BMI, body mass index.

  2. a

    P <.05.

Age    
≤60 y    
Usual-care group (n=43)0.10 (−0.70 to 0.91)−0.55 (−1.38 to 0.27)  
Intervention group (n=62)−0.33 (−0.98 to 0.33)−0.69 (−1.35 to −0.02)a−0.43 (−1.46 to 0.61)−0.13 (−1.19 to 0.92)
>60 y    
Usual-care group (n=162)−0.54 (−1.13 to 0.05)−0.51 (−1.12 to 0.09)  
Intervention group (n=143)−0.78 (−1.42 to −0.13)a−1.41 (−2.07 to −0.75)a−0.23 (−1.11 to 0.64)−0.90 (−1.80 to −0.01)a
Sex    
Males    
Usual-care group (n=115)−0.30 (−1.04 to 0.45)−0.03 (−0.79 to 0.73)  
Intervention group (n=106)−0.65 (−1.43 to 0.14)−1.36 (−2.18 to −0.55)a−0.35 (−1.44 to 0.73)−1.33 (−2.44 to −0.21)a
Females    
Usual-care group (n=90)−0.60 (−1.18 to −0.02)a−1.22 (−1.82 to −0.62)a  
Intervention group (n=99)−0.64 (−1.20 to −0.08)a−1.04 (−1.61 to −0.47)a−0.05 (−0.85 to 0.76)0.18 (−0.65 to 1.01)
Site    
Colon    
Usual-care group (n=131)−0.44 (−1.05 to 0.18)−0.46 (−1.09 to 0.18)  
Intervention group (n=145)−1.04 (−1.63 to −0.45)a−1.31 (−1.91 to −0.70)a−0.60 (−1.45 to 0.25)−0.85 (−1.73 to 0.03)
Rectum    
Usual-care group (n=74)−0.44 (−1.25 to 0.37)−0.69 (−1.50 to 0.13)  
Intervention group (n=60)0.27 (−0.64 to 1.17)−0.95 (−1.87 to −0.02)a0.70 (−0.51 to 1.92)−0.26 (−1.49 to 0.97)
BMI    
Nonobese (<30 kg/m2)    
Usual-care group (n=165)−0.47 (−1.05 to 0.10)−0.32 (−0.92 to 0.28)  
Intervention group (n=162)−0.69 (−1.28 to −0.10)a−1.42 (−2.03 to −0.81)a−0.22 (−1.05 to 0.61)−1.10 (−1.96 to −0.25)a
Obese (≥30 kg/m2)    
Usual-care group (n=39)−0.27 (−1.15 to 0.61)−1.29 (−2.17 to −0.41)a  
Intervention group (n=43)−0.50 (−1.36 to 0.36)−0.39 (−1.26 to 0.48)−0.22 (−1.56 to 1.01)0.90 (−0.34 to 2.14)
Physical activity    
Sufficient activity (≥150 min/wk)    
Usual-care group (n=24)0.77 (−0.37 to 1.92)−0.15 (−1.43 to 1.14)  
Intervention group (n=20)−0.94 (−2.03 to 0.15)−0.99 (−2.12 to 0.15)−1.71 (−3.29 to −0.13)a−0.84 (−2.57 to 0.89)
Insufficient activity (<150 min/wk)    
Usual-care group (n=181)−0.62 (−1.17 to −0.08)a−0.62 (−1.18 to −0.07)a  
Intervention group (n=180)−0.63 (−1.19 to −0.07)a−1.25 (−1.82 to −0.67)a−0.01 (−0.79 to 0.77)−0.62 (−1.42 to 0.17)

DISCUSSION

The current study presented the findings from what to our knowledge is the first randomized controlled trial to examine an intervention addressing sedentary behavior, albeit within a multiple health behavior change intervention, in survivors of cancer. Among colorectal cancer survivors, prolonged sitting time is associated with poorer prognosis[12] and achieving a healthy BMI is associated with an improved prognosis.[28] CanChange, a broad multiple health behavior change intervention, did not produce a significant between-group decrease in television viewing time, screen time, or total sedentary time at either 6 months or 12 months. However, the net difference of 0.65 hours per day (39 minutes/day) of total sedentary time may represent a clinically significant change.[29] Replacing 0.65 hours per day of sitting with light-intensity physical activity would expend an additional 45 kilocalories per day in an average adult weighing 70 kilograms (kg). If sustained on a daily basis, this small change in energy expenditure would result in a loss of 2 kg over the course of 1 year. Nevertheless, a sedentary behavior-specific intervention may be necessary to achieve more substantial, and statistically significant, reductions in sedentary time.

Despite no overall effect, several subgroup effects were noted. Participants who were older (aged > 60 years), male, and nonobese were more likely to reduce their sedentary behavior in response to the CanChange intervention. These findings suggest that some survivors of colorectal cancer may require more intensive intervention than others if substantial reductions in sedentary behavior are to be achieved across all clinical colorectal cancer survivor groups. We did note a significant difference in total sedentary behavior between the health coaching and usual-care groups at 6 months among the sufficiently active participants, suggesting that no compensatory effect occurred. However, these findings must be interpreted with caution given the small number of participants who were achieving 150 minutes per week of leisure-time physical activity.

To our knowledge, no previous study has examined the effects of a behavioral intervention on the “sitting time” of cancer survivors. A recently published randomized controlled trial in another chronic disease high-risk population provides some perspective on the current findings. A multiple health behavior change intervention delivered to Dutch adults considered to be at risk of type 2 diabetes or cardiovascular disease also resulted in nonsignificant between-group changes in sedentary behavior. The Hoorn Prevention Study randomized 622 adults to a lifestyle intervention delivered by general practice nurses (6 face-to-face consultations followed by 3 telephone follow-up sessions over a 6-month period) or a control group (provided with healthy lifestyle brochures). At 12 months of follow-up (the primary endpoint adopted by the CanChange trial), there was no statistically significant between-group effect for total or domain-specific sedentary behavior observed. There was a net difference of 15 minutes of total sedentary behavior per day between the groups at this time point.[30]

To the best of our knowledge, the majority of the sedentary behavior intervention trials conducted to date have been performed within the workplace setting. Until recently, the primary objective of these workplace interventions was to increase physical activity levels. A review of sedentary behavior interventions in the workplace demonstrated that interventions aiming to increase physical activity in the workplace did not result in significant changes in sedentary behavior.[31] However several studies have recently demonstrated significant decreases in total sitting time, or decreases in periods of prolonged sitting, in office environments by specifically focusing on sedentary behavior.[32-36] The most successful and sustainable sedentary behavior change intervention used an ecological approach (addressing organizational-level, environmental-level, and individual-level factors) in the design and implementation of the intervention.[34] Although interventions for the workplace setting are not directly focused on cancer survivors, lessons from their success can be drawn. Future interventions for cancer survivors may benefit from attempting to address features such as the social and physical (home) environment, as well as individual-level factors.

Key strengths of the current study include the use of a randomized controlled design, a theory-based multiple behavior change intervention, and assessment of health behaviors at the completion of the intervention (6 months) and after a period of no-intervention contact (12 months). The primary limitation of the current study was the use of a self-reported measure of sedentary behavior. The measure had modest validity and reliability,[22] but its psychometric properties were similar to those of other sedentary behavior measures available at the time of the study. Future studies would benefit from the use of a more accurate and comprehensive assessment of sedentary behavior, such as previous-day recalls or objective measures of sedentary time.[37, 38]

In the current study, we observed improvements in reported health behaviors in the control group, a phenomenon that has been reported in other trials.[39-41] Reasons for the reduction in sedentary behavior in the usual-care group are unclear, although it is possible that survivors of colorectal cancer who consented to participate in the CanChange trial were highly motivated to improve their health behaviors. Selection bias is an inherent limitation in randomized controlled trials. Individuals who are highly motivated to improve their health and to participate in studies may not be representative of the whole population of adults diagnosed with colorectal cancer in Australia. Therefore, we cannot confidently conclude that the intervention effects noted within the current trial can be generalized to the broader population of colorectal cancer survivors. Furthermore, given that this trial was not originally powered to detect subgroup differences, and the potential for false-positive findings from multiple subgroup analyses, definitive trials testing interventions in targeted subgroups with a stronger focus on sedentary behavior are warranted.

Many colorectal cancer survivors are highly motivated to adopt behavioral changes that will enhance their quantity and quality of life. Reducing sedentary behavior holds promise as an additional behavioral strategy for improving prognosis and long-term health outcomes among survivors of colorectal cancer7; efficacious sedentary behavior change interventions are therefore essential. The CanChange multiple health behavior change intervention achieved a potentially meaningful, but nonsignificant, between-group difference in total sedentary time in the overall sample, and significant differences in some subgroup analyses. A stronger focus on reducing sedentary behavior in future interventions may lead to greater reductions in sitting time among colorectal cancer survivors. Future research should also consider the timing of the intervention delivery because participants may be more receptive to lifestyle changes at different points across the survivorship trajectory.[42]

FUNDING SUPPORT

Funding received from the Australian government through Cancer Australia.

CONFLICT OF INTEREST DISCLOSURES

Dr. Lynch is supported by a Public Health Training Fellowship from the National Health and Medical Research Council (586727) and the Victorian government's Operational Infrastructure Support Program. Dr. Courneya is supported by the Canada Research Chairs Program.

Ancillary