Narrative Review of Web-based Healthy Lifestyle Interventions for Cancer Survivors

This narrative review of web-delivered weight management, diet quality, and physical activity interventions for cancer survivors relies on a systematic search of PubMed, Psych Info, and EBSCOhost which identified 19 unique web-delivered lifestyle interventions for cancer survivors. The sample sizes for these studies ranged from 11–492. Intervention duration ranged from 1–12 months; however, most interventions were 6–12 weeks in length. Ten studies were randomized controlled trials (RCTs), two were two-arm quasi RCTs, and seven employed a single-arm pre/post-test design. Many (N= 15) of the interventions were well-grounded in behavioral theory, which may have led to favorable behavior change. Most studies (15-of-19) targeted and reported increases in physical activity, while only a few targeted and reported improvements in diet quality (36.9% and 15.8%, respectively) and weight management (26.3% and 10.5%, respectively). A notable limitation was that most studies were conducted among populations that were primarily White and female. Future directions for Internet-based lifestyle interventions for cancer survivors include increasing: (a) focus on multiple behavior change, (b) representation of male and minority populations to improve generalizability of findings, (c) extended intervention duration and follow-up to evaluate long-term efficacy of web-based lifestyle interventions, and (d) sample size to allow for adequate statistical power.


Introduction
By the year 2029, there will be an estimated 21.7 million cancer survivors [1]; a population increasingly growing due to the aging population, improved early detection, and modern therapeutics and treatment modalities [1]. However, cancer survivors represent a vulnerable population characterized by high rates of obesity, physical inactivity, poor diets, and related chronic conditions [2][3][4]. Face-to-face lifestyle interventions have shown promise for improving the quality-of-life and reducing adverse health outcomes among cancer survivors [5][6][7][8]. However, these approaches are relatively costly and thus have limited reach to the survivors who need them most [8]. Web-delivered lifestyle interventions have already shown success in other populations [9,10] and have the potential to reach large numbers of cancer survivors at relatively low cost. Moreover, web-based strategies can overcome obstacles cited by cancer survivors by not requiring transportation or clinic visits [8]. In fact, they can be accessed 24/7 from any Internet-enabled device [11][12][13]. Past studies indicate that cancer survivors are already using the Internet to obtain healthy lifestyle information [14] and report a lack of credible online resources for healthy lifestyle behavior change information [15]. Thus, there is an evident interest and need for web-based lifestyle intervention in this patient population.
Findings-to-date from studies in this area have been encouraging. Prior related reviews on non-face-to-face lifestyle interventions (including only three web-based interventions) [16], eHealth physical activity interventions [17], and digital diet and physical activity interventions for cancer survivors [18] all found the number of health behavior change interventions are increasing. It is important to note that digital and eHealth interventions are distinct from web-based interventions. Web-based interventions require an Internet connection for program delivery, whereas eHealth interventions use the Internet specifically to provide healthcare services, while digital programs do not require the Internet for intervention delivery (e.g., Gaming systems like Wii can be operated by inserting a game into a console without an Internet connection). Nevertheless, there has yet to be a review of web-based physical activity and diet interventions (as both behaviors are critical to maintaining a healthy weight) for cancer survivors. Therefore, there is a need for a review of web-based lifestyle intervention research in cancer survivors to help identify the next steps and move this field forward, given the critical public health implications and rapidly evolving nature of the Internet.

Study Selection
PubMed, Psych Info, and EBSCOhost were systematically searched for articles published on web-based nutrition, physical activity, and/or weight management interventions for adult cancer survivors. The research intersected terms for physical activity (exercise), weight (weight loss, weight management, weight change, weight reduction), nutrition (diet), Internet (web, web-based), cancer (cancer survivorship, tumor, neoplasm) and interventions (programs). Duplicate papers, papers not written in English, those containing non-human and non-adult samples were manually removed through the study screening process. Included studies met the following criteria: web-or Internet-based programs requiring an Internet connection to access intervention materials (e.g., sending or receiving an email, visiting a website or accessing a mobile application) and targeting weight management, physical activity, and/or nutrition in adult cancer survivors diagnosed with any cancer type during or post-treatment.
Studies were excluded for the following reasons: a) study was not exclusively for cancer survivors; b) non-web or Internet-based studies; c) not in English; d) only available as an abstract; d) overlapping publications; e) protocol papers, and f) review papers. Intervention studies with quasi-experimental and pre-/ post-test research designs were included as this is an emerging field. Those early studies may help shed light on future directions and provide a holistic view of current research. Two researchers independently identified and reviewed studies potentially meeting eligibility criteria. First, the titles and abstracts were reviewed, and duplicate articles were removed. For studies passing the initial screening, the article's full text was reviewed. A data extraction form was used to record pertinent information on each study consistently. The form included sections on the papers' authors, country of origin, publication year, study design, cancer type, intervention activities, number of participants, intervention duration, and behavioral outcomes. Any discrepancies were resolved by consensus or through discussion with a third researcher.

Results
See Figure 1 for the PRISMA flow diagram of the study selection process. The search strategy initially identified 862 records; 19 unique web-based lifestyle interventions for cancer survivors met the inclusion criteria and were included in this review. The included articles were published between 2011 and 2018. Table 1 provides a summary of each of the studies included in the review.

Discussion
Internet-based approaches to promote behavioral change in cancer survivors appear promising. All the included studies targeted physical activity, and most were associated with increases in physical activity among cancer survivors. Few interventions targeted weight management and diet quality. However, an estimated 20% of cancer cases and 30% of cancer deaths are attributed to the combined effects of an unhealthy diet, excess body weight, and physical inactivity [40,41]. Lifestyle behaviors are modifiable cancer risk factors that can be addressed to improve the quality and quantity of life among cancer survivors [42]. Therefore, future interventions should target a combination of physical activity, weight management, and diet quality as cancer survivors often require multiple behavior changes to improve their quality of life [43,44]. Behavior change theory has been shown to improve the effectiveness of health behavior change interventions [45][46][47]. Yet, the descriptions of the theoretical framework used for intervention development varied in detail. Most studies explained how the theory was used for program development (e.g., theoretical constructs were targeted by specific intervention strategies and incorporated into the assessment), but some did not. Future web-delivered lifestyle interventions for cancer survivors should continue using theoretical frameworks for program development and explicitly report how theory is integrated.
Studies included in this review generally focused on short-term improvements in health behavior and tended to forgo following up post-intervention to see if these changes were maintained. While short-term change is favorable, long-term health behavior modifications have lasting health implications (e.g., reduced cancer and recurrence risk). Maintaining healthy lifestyle behaviors is often a challenge as people return to unhealthy lifestyle behaviors [48]. Nevertheless, lengthier interventions and follow-up periods are needed to determine whether web-delivered lifestyle interventions facilitate long-term behavior change. Small sample sizes also were a major limitation of many of the studies that were reviewed. Among the RCTs, many did not appear to have appropriate sample sizes for statistical power. Quasi-RCTs and single-arm pre-/post-test study designs were included in this review as they highlight ongoing research in the field. However, these studies were also typically underpowered. Therefore, future research should include larger sample sizes for greater statistical power to detect intervention effects, especially among segments of cancer survivors.
Lifestyle interventions for cancer survivors have been predominately for cancer-specific populations [16]. A prior review of broad-reach (comprised of telephone, print, and web intervention) lifestyle interventions for cancer survivors found that most programs were conducted with single cancer populations (i.e., exclusively for breast or prostate cancer survivors [16]. However, unhealthy lifestyle behaviors are present and problematic across all cancer types and require appropriate interventions [1]. Encouragingly, this review found that web-delivered lifestyle intervention studies included more "diagnosis diverse" (cancer survivors of more than cancer-type) populations than tend to be reported for face-to-face clinic-based interventions. Of the 19 studies included in the review, 12 studies included diagnosis diverse populations with three studies targeting two populations of cancer survivors (e.g., breast and endometrial; colorectal and prostate; and testicular and breast) and two studies targeting three populations of cancer survivors (e.g., breast, colorectal, and prostate). Therefore, only seven studies targeted single diagnosis populations (e.g., breast and prostate cancer survivors). Despite cancer diagnosis diversity among participants, populations were still predominately female. For example, Puszkiewicz et al. included breast, prostate, and colorectal cancer survivors and the population was 82% female. Therefore, future studies should target diagnosis diverse male cancer survivors. Studies with diagnosis diverse populations and male representation increase the generalizability of findings among cancer survivor groups, as does minority representation.
There is still a lack of participation from ethnic and racial minorities in healthy lifestyle cancer research. Several studies (N=6) did not report the racial/ethnic distribution of their study samples and of the studies that did samples were predominately White, except for one study among African American breast cancer survivors [26]. African Americans are especially vulnerable to poor health outcomes and are disproportionately affected by many health conditions associated with poor diet quality, excess body weight, and physical inactivity [49]. More studies should specifically target racial and ethnic minorities, especially African Americans, to provide much-needed interventions and eliminate related cancer disparities.

Conclusion
Web-delivered lifestyle interventions provide a means for delivering health behavior change programs to meet the growing needs of cancer survivors. Overall, the literature suggests that web-delivered lifestyle interventions have the potential to promote healthy lifestyle behavior among cancer survivors, especially increasing physical activity. However, future lifestyle interventions should address methodological issues outlined in this article (e.g., multiple behavior change, targeting minority and male survivors, post-intervention follow-up periods, and larger sample sizes).