Psychosocial Determinants of Lifestyle Change after a Cancer Diagnosis: A Systematic Review of the Literature

Simple Summary Although cancer survivors may experience health benefits from favorable lifestyle changes, many cancer survivors do not adhere to lifestyle recommendations or make favorable lifestyle changes after cancer diagnosis. This systematic review of the literature aimed to provide an overview of the scientific literature on sociodemographic, psychological and social determinants that may facilitate or hamper lifestyle change after the diagnosis cancer. It provides a structured overview of the large variety of determinants of changes in different lifestyle behaviors (physical activity, diet, smoking, alcohol, sun protection, and multiple lifestyle behaviors) derived from the 123 included papers (71 quantitative and 52 qualitative). Findings demonstrate the important role of oncology healthcare professionals in promoting healthy lifestyle changes in cancer survivors and inform researchers and healthcare professionals about the methods and strategies they can use to promote healthy lifestyle changes in cancer survivors. Abstract The aim of this study is to provide a systematic overview of the scientific literature on sociodemographic, psychological and social determinants that may facilitate or hamper lifestyle change after the diagnosis cancer. Four databases (PubMed, PsychINFO, Cumulative Index to Nursing and Allied Health Literature (CINAHL), and Web of Science) were searched for relevant papers up to October 2021. Of the 9586 references yielded by the literature search, 123 papers were included: 71 quantitative and 52 qualitative papers. Findings showed a large variety of determinants influencing lifestyle change after cancer diagnosis, with differences between lifestyle behaviors (physical activity, diet, smoking, alcohol, sun protection, and multiple lifestyle behaviors) and findings from quantitative vs. qualitative studies. Findings demonstrate the important role of oncology healthcare professionals in promoting healthy lifestyle changes in cancer survivors. In addition, findings inform researchers involved in the development of health promotion programs about the methods and strategies they can use to promote healthy lifestyle changes in cancer survivors. Favorable lifestyle changes are expected to have beneficial effects on cancer risk and overall health in cancer survivors.


Introduction
A large body of evidence has demonstrated that lifestyle not only influences the risk of developing cancer [1] but also the risk of cancer recurrence, comorbidities such as cardiovascular disease and type II diabetes mellitus, and mortality [2][3][4][5][6]. Moreover, lifestyle has been associated with several biological mechanisms, such as inflammation and Natural Killer cell function, that may impact health-related outcomes [7][8][9][10]. Favorable lifestyle changes, such as increasing physical activity or smoking cessation, may optimize these health outcomes and increase health-related quality of life among cancer survivors (i.e., individuals who are living with a diagnosis of cancer, including those who have recovered from the disease [1]) [11][12][13][14][15][16]. In accordance, lifestyle and body weight recommendations have been issued for cancer survivors, such as the recommendations from the World Cancer Research Fund (WCRF) [1,17]. Despite the potential health benefits, many cancer survivors do not adhere to these recommendations and do not make favorable lifestyle changes after diagnosis [18][19][20][21][22][23]. The reason for this is likely to be complex and multifactorial.
Knowledge on determinants that enhance lifestyle changes (i.e., facilitators) and determinants that limit lifestyle changes (i.e., barriers) in cancer survivors is needed to be able to identify what techniques and strategies can be used to achieve lifestyle changes in this specific patient population. It is important to use behavior change techniques and strategies matching cancer survivor specific determinants, as these are likely to require a (partly) different approach as opposed to other patient populations or the general population. Park & Gaffey (2007) provided an overview of psychosocial determinants of lifestyle change after a cancer diagnosis [24]. Their integrative review included the results of 30 quantitative studies examining relationships among psychosocial factors and lifestyle change in cancer survivors. They concluded that findings of the included studies were inconsistent and that their ability to draw conclusions was limited, predominantly due to mostly cross-sectional study designs and the heterogeneity between the included studies.
To extend the existing literature, this study builds on the review of Park & Gaffey (2007) [24] with an updated, extended, systematic literature search, structured per lifestyle behavior, and additionally including qualitative research. Using both quantitative and qualitative research methods to gain knowledge on determinants, capitalizes the strengths of both research methods [25]. The aim of this study is to provide a systematic overview of the scientific literature on sociodemographic, psychological and social determinants that may facilitate or hamper lifestyle change after the diagnosis cancer. Data on sociodemographic determinants (such as age, gender, educational level, and marital status) may provide insight into which cancer survivors specifically should be targeted to promote lifestyle changes. Data on psychosocial determinants, both at the inter-individual level (determinants at the between-person level, such as social support) and intra-individual level (determinants at the within-person level, such as self-efficacy), provides insight into which modifiable determinants should be targeted for change and informs about what type of techniques or strategies can be used to positively influence these modifiable determinants.

Materials and Methods
This systematic review of the literature was conducted in line with PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines and was submitted to PROSPERO (International prospective register of systematic reviews; ID313277).

Literature Search
A systematic review of the literature up to the 20 October 2021 was conducted. A total of four databases were searched for relevant papers: (PubMed, PsychINFO, Web of Science, and CINAHL (Cumulative Index to Nursing and Allied Health Literature). A combination of search terms from the following concepts were used: Cancer survivors AND one of the lifestyle behaviors (lifestyle, physical activity, diet, smoking, alcohol consumption, OR sun protection) AND change AND psychosocial (psychological OR social). The complete list of search terms used associated with each concept included in the search is provided in Table 1. When performing the search in the databases a filter for language was applied, including only articles in the English language. Articles retrieved from the database searches were exported to a reference library (EndNote) and combined into one database, in which duplicates were deleted.

Selection Procedure
First, two researchers (IN and MH) simultaneously screened and labelled 10% of the retrieved articles based on title and abstract. Inconsistencies in labelling were resolved by discussion until consensus was achieved. Second, the remaining articles (90%) were divided among these two researchers and were screened and labeled based on title and abstract. After, the full-texts of the articles that were not excluded based on title and abstract were divided among three researchers (IN, CH and MH) and were read and labelled independently to select eligible full-texts. Inconsistencies between the researchers with regard to whether an article should have been included or not were discussed until consensus on inclusion or exclusion was achieved.
Both observational and intervention studies describing psychosocial determinants of change in physical activity, dietary intake, smoking, alcohol consumption, and sun protection among survivors of any type of cancer and any time since diagnosis were included. Articles on other outcome measures (e.g., changes in sleep or weight loss) were excluded. Furthermore, an article was excluded when it described non-human research, when the described study was not original research (e.g., a review article), when the study population did not (only) consist of cancer survivors, when the study did not describe change in one of the lifestyle behaviors of interest, and when the article did not involve psychosocial determinants.

Data Extraction
The following data were extracted and described separately for each included article: first author and year of publication, country, study design, sample characteristics (sample size, type(s) of cancer, mean age with standard deviation (SD), percentage of female participants, mean time since diagnosis or treatment, and mean baseline Body Mass Index (BMI), psychosocial and lifestyle behavior measurements including measurement instruments, and the findings with regard to psychosocial determinants of lifestyle change. Psychosocial variables were categorized into socio-demographic, inter-individual, and intra-individual determinants.

Results
In Figure 1, a flow diagram is depicted of inclusion and exclusion of publications derived from the database searches, including reasons for exclusion. In total, the database searches yielded 9586 references. After removal of 2979 duplicates, 6607 titles and abstracts were assessed for eligibility. Of the 176 full-texts that were screened, 123 papers were included: 71 quantitative papers and 52 qualitative papers. See Tables S1-S10 for an overview of the characteristics and findings of the included quantitative and qualitative studies.

Psychosocial Determinants
An overview of the psychosocial determinants of lifestyle changes retrieved from the included quantitative studies is presented in Table 2. See Table 3 for an overview of the psychosocial determinants of lifestyle changes retrieved from the included qualitative studies. Below, both quantitative and qualitative findings on psychosocial determinants of changes in lifestyle are presented per lifestyle behavior.

Sun Protection (n = 2)
Cancer-related avoidance NS [23] Fear of exercise NS [29] Fear of recurrence        Not wanting to bother the host with dietary restrictions [11,40] Perceiving smoking as a social norm and as a tool for communication and connecting with friends [41] Feeling impolite or embarrassed to reject food prepared by others/a cigarette from a friend [40] [41] Dilemma between staying on a healthy diet and maintaining harmony with others [40] Residing with other smokers [5] Social pressure (e.g., pressure to stop smoking from relatives) Timing of the intervention (during radiotherapy) [11] Unfavorable lifestyle and lack of lifestyle change in social environment [14,29] Difficulties in shopping for food [11] Specific social events [11,40] [14] Unexpected (major) life events (e.g., serious illness, death) [22] [29] Belief that weight loss is a positive health outcome of cancer [34]

Changes in Physical Activity
In total, 71 of the included studies described psychosocial determinants of changes in physical activity in cancer survivors (45 quantitative studies and 26 qualitative studies).
Eight out of the 12 quantitative studies assessing socio-demographic determinants assessed educational level as a determinant of changes in physical activity [28,33,44,76,78,80,84,87], of which five did not find a significant association between educational level and changes in physical activity [28,33,76,78,80]. The three studies that did find a significant association showed mixed results (see Table 3) [44,84,87]. In the qualitative studies, educational level was not mentioned.
Five quantitative studies assessed employment status as potential determinant of changes in physical activity [28,33,78,80,87]. Four out of these five studies did not find a significant association. The one study that did, found increasing sedentary time to be higher in participants working fulltime [33]. In the qualitative studies work-related factors (e.g., resuming work, working full-time) were mentioned as barriers to favorable changes in physical activity [99,113,145].
Other socio-demographic determinants that were found to be significantly associated with changes in physical activity were job position [80], social class [85], and income [44]. In two qualitative studies [100,109,119,132,146], financial constraints were mentioned as a barrier to changes in physical activity (e.g., the cost of attending physical activity facilities). On the other hand, affordability, was mentioned as a facilitator [100,109].
Marital status [28,78,80,85,87], race/ethnicity [28,44], and years of education [94] were not found to be associated with changes in physical activity. Gender was assessed as a potential determinant in two quantitative studies that found mixed results [33,39]. See Table 2. Gender was not specifically mentioned as a determinant in the qualitative studies.
In the qualitative studies, poor weather conditions was frequently mentioned as a barrier to changes in physical activity [97][98][99]109,[112][113][114]119,129,131,146,147]. Environmental factors, such as poor infrastructure, geographical isolation, and lack of footpaths, were also mentioned as barriers in qualitative studies [114,129,132], whereas a pleasant local physical activity environment was mentioned as a facilitator of physical activity changes [132]. See Table 3.
See Table 2 for an overview of the results of the few quantitative studies assessing interindividual determinants other than social support (role models [43,81], social modeling [38]).
Cognitive and behavioral factors were also mentioned as determinants of changes in physical activity. Of the four studies examining cognitive and behavioral processes [33,34,37,40], three studies found a significant positive association with favorable changes in physical activity [34,37,40] and two studies found conflicting associations [33,34]. Two quantitative studies examining action planning both found a significant association with changes in physical activity [77,96], but not for coping planning [96]. In the qualitative studies, action planning and goal setting was mentioned as a facilitator [101,114,132,146]. In one quantitative study, goal setting was associated with changes in physical activity [74]. Also, (self-)monitoring and feedback on behavior [97,98,101,132,137,[142][143][144][145] was frequently mentioned as a facilitator in the qualitative studies. Two of the three quantitative studies examining decisional balance found it not to be significantly associated with changes in physical activity [34,37], whereas the other study found decisional balance to be associated with physical activity adoption, but not maintenance [34]. Also, a higher stage of change (a higher readiness to change) was found to be a significant positive predictor of change in physical activity [76].
Determinants related to motivation for physical activity changes were examined in four quantitative studies. Two quantitative studies assessing motivation found conflicting results [74,75]. One study assessing motivational regulations found changes in self-determined motivation to be positively related to changes in physical activity [65]. Another quantitative study examining motivational processes found that perceived opportunity was a significant mediator of exercise behavior [69]. In the qualitative studies, personal and/or internal motivation was mentioned as a facilitator [101,113,137,141,143], whereas lack of motivation [98,99,101,109,112,114,119,[129][130][131][132][145][146][147] was mentioned as a barrier for changes in physical activity.
The included studies reported on the relation between perceptions or expectations and changes in physical activity. Four quantitative studies assessed perceived barriers, of which two found no association with changes in physical activity [73,74], and the other two showed mixed results [42,94]. Six quantitative studies examined outcome expectations as a determinant of change in physical activity [27,30,42,43,74,94], of which two found a positive association [74,94]. Perceptions of physical activity improving quality of life and overall survival was found to be associated with increased physical activity [67], while exercise beliefs of negative impact of exercise on cancer was found to be associated with decreased physical activity [73]. In the qualitative studies, perceived or anticipated benefits of lifestyle change (e.g., to improve health, wellbeing, reduce symptoms, improving treatment efficacy & cancer prognosis) were mentioned as facilitators [98,101,111,119,132,144,147].
Furthermore, experienced benefits from physical activity (e.g., improving mental wellbeing, processing negative thoughts and feelings) [97,98,100,101,110,111,114,119,120,129,132,136,137,[141][142][143][144][145][146][147] were frequently mentioned as facilitators in the qualitative studies. Another frequently mentioned facilitator of physical activity changes mentioned in the qualitative studies was enjoyment of being physically active [98,109,113,119,120,141,146], whereas lack of enjoyment of physical activity [98,99,132,137,143] and not being the sporty type [99,101,114,131,145,146] were mentioned as barriers. One of the two quantitative studies on physical activity enjoyment found no significant association [43], the other found that an increase in physical activity enjoyment significantly predicted physical activity at post-intervention [82]. Another frequently mentioned facilitator in the qualitative studies was the perception that being more physically active was experienced as a way of being able to do something and re-gain control over their lives [97,98,100,113,120,136,144,146].
Four quantitative studies assessed fatigue as potential determinant [26,64,68,94]. Three studies found fatigue to be a significant determinant of changes in physical activity [26,64,68], with less fatigue being associated with increased physical activity [26], fatigue being a significant predictor of physical activity maintenance [68], and participants with higher levels of fatigue were less likely to remain consistently sufficiently active [64].

Dietary Changes
30 studies reported on psychosocial determinants of dietary changes (21 quantitative studies and nine qualitative studies).

Socio-Demographic Determinants
Nine studies assessed socio-demographic determinants of dietary changes in cancer survivors [61,70,76,78,80,87,[90][91][92]. Seven out of those nine studies assessed age as a determinant of lifestyle changes [61,70,78,80,87,90,91]. Of the four studies that did find an association between age and dietary changes [70,78,90,91], three found that younger cancer survivors were more likely to make favorable dietary changes [70,90,91] and one found that older cancer survivors were more likely to make favorable lifestyle changes [78]. Ageing was not mentioned as a determinant of dietary changes in the qualitative studies.
Eight studies assessed educational level as a potential determinant of dietary changes [61,70,76,78,80,87,91,92]. Six out of these eight studies did not find a statistically significant association between educational level and dietary changes [70,76,78,80,87,91], while two studies found that a higher level of education was associated with making favorable dietary changes [61,92]. All four quantitative studies assessing associations between marital status and dietary changes found no statistically significant associations [78,80,87,91].

Inter-Individual Determinants
Although marital status specifically was not mentioned in the qualitative studies, social support from family, friends, and health care professionals was frequently mentioned as a facilitator of dietary changes [102][103][104]122,123,138,139]. In three of the quantitative studies, social support was assessed as a potential determinant of dietary changes [26,35,44]. Two out of these three studies found that social support determined favorable dietary changes [35,44]. In the qualitative studies, many other inter-individual determinants were reported (see Table 3), such as lack of information or advice from health-care professionals as a barrier to dietary changes [102,103,138].
Of the five quantitative studies assessing the association between self-efficacy and dietary changes [35,47,48,50,76], three found statistically significant associations indicating that higher self-efficacy was associated with favorable dietary changes [35,47,48]. In the qualitative studies, self-efficacy was not mentioned as a determinant of dietary changes.
Four quantitative studies examined stress-related variables: stressful life events [26,70], contemporary life stress [36], psychological distress at diagnosis [70], and cancer-related stress [49]. Although one study found that a greater number of stressful events in the five years preceding diagnosis was associated with initiating dietary change [70], other studies found no statistically significant association between stressful life events [26] or contemporary life stress [36] and dietary changes. One study found that higher initial psychological distress at diagnosis was associated with initiating dietary change [70]. Another study found that cancer-related stress was a barrier to fruit and vegetable consumption around the diagnosis, but facilitated positive dietary changes by the end of the first year after diagnosis [49]. In the qualitative studies, stress-related variables were not specifically mentioned as determinants of dietary changes.
See Table 2 for the results on the intra-individual determinants of dietary changes examined by one or two quantitative studies, such as perceived barriers, health-related quality of life, fear of recurrence, stage of change illness representations, perceived behavioural control, dispositional optimism, and cancer coping style [26,31,44,46,48,50,76,87,89,92,102,103,121,123,139].
Frequently mentioned barriers to dietary changes in the qualitative studies that were not assessed in the quantitative studies include perceived/anticipated benefits of lifestyle change (e.g., to improve health, wellbeing, reduce symptoms, improving treatment efficacy & cancer prognosis) [102,115,121,138] and lifestyle change as active coping strategy: doing something to gain a sense of control [102,103,115,121,138]. See Table 3 for an overview of the determinants of dietary change mentioned in the qualitative studies.

Changes in Smoking Behavior
16 studies described psychosocial determinants of changes in smoking behavior, of which 12 were quantitative and four were qualitative studies.

Sociodemographic Determinants
Eight quantitative studies assessed socio-demographic determinants of changes in smoking behavior in cancer survivors [52,53,55,56,59,60,93,95]. Seven out of those eight studies assessed age as a determinant of changes in smoking behavior [52,53,55,56,59,60,95]. Four out of those seven studies did not find a significant association between age and changes in smoking behavior [53,55,59,60]. The other three studies that did find an association between age and changes in smoking behavior found that older participants were more likely to have been abstinent from smoking [52,56,95]. Age was not mentioned as a determinant of changes in smoking behavior in the qualitative studies.
Six out of the eight quantitative studies assessed educational level as a determinant of change in smoking behavior [52,53,59,60,93,95]. Five out of these six studies did not find a statistically significant association between educational level and changes in smoking [52,59,60,93,95]. The one study that did find a significant association between educational level and changes in smoking found that long-term cessation rates were lower among those with lower educational levels [53]. Educational level was not mentioned as a determinant of changes in smoking behavior in the qualitative studies.
Of the five quantitative studies assessing marital status as a determinant of change in smoking behavior [52,53,59,93,95], one found a marginally significant association between marital status and changes in smoking behavior, with married participants yielding higher abstinence rates in the intervention group [59]. Marital status was not explicitly mentioned as a determinant in the qualitative studies.
Six quantitative studies assessed gender as a predictor of changes in smoking behavior [53,55,56,59,60,95]. Whereas five of these studies did not find a significant association [53,55,59,60,95], one study found that participants were more likely to have been abstinent at one of the follow-up measurements if they were male [56]. The four quantitative studies assessing race [53,55,59,60] found no statistically significant associations.
The two quantitative studies assessing income [53,95] as potential determinant of changes in smoking behavior found no statistically significant associations. In the qualitative studies, lack of work (e.g., being unemployed or not able to work after cancer diagnosis) was mentioned as a barrier to smoking cessation [133]. Also, affordability and smoking cessation saving money were mentioned as facilitators of smoking cessation in the qualitative studies [105,140].
Two quantitative studies examined second-hand smoke exposure at home [52,95], of which one study found that being exposed to second-hand smoking at home was significantly associated with being indecisive for abstinence [95]. The other study did find a significant association between having household members that smoke and continued smoking univariately, which only remained marginally significant when examined multivariably [52].

Inter-Individual Determinants
Two quantitative studies examined inter-individual determinants of changes in smoking behavior [55,71]. One study did not find social support to be a significant predictor of smoking cessation [71], whereas the other study did find significant differences between continuous abstainers and participants that relapsed in some, but not all, supportive behaviors [55]. One study assessed social smoking environment as possible determinant and found that participants were more likely to quit smoking if they had a spouse who did not smoke, and fewer peers who smoked [71].
In the qualitative studies, social support (e.g., from partners and family members) [124,133,140], advice or support from health care professionals [124,133], and the social unacceptability of smoking [105,124] were mentioned as facilitators of favorable changes in smoking behavior. Lack of discussion about lifestyle with health care professionals [105,124] was mentioned as a barrier to favorable changes in smoking behavior. See Table 3 for an overview of all determinants at the inter-individual level retrieved from the qualitative studies.
Of the six quantitative studies assessing depression as a determinant of changes in smoking behavior, three did not find a significant association [55,59,95]. The other three studies did find depression to be a significant predictor of changes in smoking behavior, with depression being associated with continued smoking [52], relapse after quitting [53], and lower abstinence rates [57]. One of the three quantitative studies examining anxiety as a potential determinant of change in smoking behavior found that lower levels of anxiety significantly predicted abstinence [60]. The other two studies did not find significant associations between anxiety and change in smoking behavior [55,95]. Four quantitative studies assessed whether self-efficacy was a determinant of changes in smoking behavior [53,55,59,95], of which three studies found that long-term cessation [53] and perseverance for abstinence [95] were less likely among participants with lower selfefficacy, and that relapsers expressed significantly lower levels of confidence in their ability to stay off cigarettes [55].
Four quantitative studies assessed stages of change [53][54][55]58]. Of the three studies that found significant associations, one found a relationship between stage of change and long term smoking status [53], one study found that participants with a higher stage of change were more likely to quit smoking [58], and one study found stage of change to significantly differentiate between continuous abstainers and relapsers, with the higher the stage of change, the less likely the patient was to relapse [55].
Stage of change, self-efficacy, and risk perception were not mentioned as a determinant in the qualitative studies. In contrast, lack of knowledge and limited perceptions on smoking cessation and health consequences [105,140], not perceiving any benefits of smoking cessation [140], and not being too concerned about effects of smoking [133] were mentioned as barriers to favorable changes in smoking behavior in the qualitative studies. See Table 3 for an overview of the intra-individual determinants mentioned in the qualitative studies.

Changes in Alcohol Consumption
Four quantitative studies reported on determinants of changes in alcohol consumption.

Socio-Demographic Determinants
Three quantitative studies assessed socio-demographic determinants of changes in alcohol consumption in cancer survivors [61,85,93]. The two studies assessing educational level as a potential determinant of change in alcohol consumption found mixed results [61,93]. The studies assessing marital status [85,93], age [61,85], and social class [85] as potential determinants of change in alcohol consumption did not find significant associations.

Inter-Individual Determinants
The only quantitative study assessing an inter-individual determinant of changes in alcohol consumption found that social support was not a significantly associated with changes in alcohol consumption [61].

Intra-Individual Determinants
Two studies assessed intra-individual determinants of changes in alcohol consumption [31,61]. Higher fear of cancer recurrence and higher emotional distress were found to be significantly associated with increased alcohol consumption [31]. Depressive symptoms and dispositional optimism were not found to be significantly associated with changes in alcohol consumption [61].

Changes in Multiple Health Behaviors
17 studies reported on psychosocial determinants of changes in multiple lifestyle behaviors (13 qualitative papers) or a lifestyle score (four quantitative papers).

Socio-Demographic Determinants
Three of the four quantitative studies assessed socio-demographic determinants of changes in lifestyle scores consisting of a combination of multiple health behaviors [32,62,79]. All three assessed age as a determinant of change and found no significant associations with any of the lifestyle scores [32,62,79].
In one qualitative study, ageing was mentioned both as barrier (e.g., viewing themselves as too old for playing sports) and facilitator (e.g., heightened awareness of susceptibility to illness due to ageing) of lifestyle change [106]. Of the three quantitative studies assessing educational level as a potential determinant, two found no significant associations with change in lifestyle behaviors (sleep, diet, exercise, and stress management) or change in substance use (alcohol and smoking) [79], or change in diet or physical activity [62]. The other study found participants with a higher level of education to be more likely to make positive changes in physical activity or diet [32]. One study assessed gender as a determinant and found female gender to be significantly related to less positive change in substance use (smoking and alcohol consumption), but not to be related to change in lifestyle behavior (sleep, diet, physical activity, and stress management) [79]. Other socio-demographic determinants that were not found to be significantly associated with lifestyle behavior changes were marital status [62,79], employment [79], income [79], and race [32,62]. See Table 2. In the qualitative studies, poor weather conditions [107,108,125], financial constraints [106,108,118] and environmental factors (such as poor infrastructure) [108] were mentioned as barriers to lifestyle changes, while environmental factors (e.g., proper infrastructure) [108] and good weather [108] were mentioned as facilitators.

Inter-Individual Determinants
Only one quantitative study assessed inter-individual determinants of changes in lifestyle behaviors [62]. This study found that social support was a significant predictor of positive behavior change (physical activity and diet), whereas no significant associations were found with social constraints [62]. In many qualitative studies, social support from partners and family members [106][107][108]117,118,125,126,135] and advice or support from health-care professionals [106,117,125,126,128,135] were mentioned as facilitators for lifestyle changes, whereas lack of information or advice from health care professionals [106,108,118,126,135,148], poor support and understanding from family members [135], and living alone or not having a partner [107,117,125,127] were mentioned as barriers of lifestyle change. See Table 3 for an overview of all reported determinants of lifestyle change in the qualitative studies.

Intra-Individual Determinants
Three quantitative studies assessed intra-individual determinants of lifestyle change [62,63,79]. Two of those three studies examined cancer-related (dis)stress as a potential determinant. One study did not find cancer-related stress to be associated with changes in lifestyle behavior [79], whereas the other study examined two subscales of cancer-related distress (intrusions and avoidance) and found only cancer-related intrusions to be a significant predictor of positive behavior change [62]. In addition, one study found that an increase in anxiety symptoms was related to greater odds of reporting an unhealthy lifestyle (physical activity, diet, BMI, alcohol and tobacco consumption) [63]. Other intra-individual determinants found not to be significant determinants of changes in lifestyle behaviors were depression [63], and traumatic stressor response [62]. See Table 2.
In the qualitative studies, concerns or fears related to symptoms (e.g., colostomy bag leakage and accidents) [107,134], coping with (emotional dis)stress through unhealthy behaviors [126,127,135], and psychological complaints such as low mood, depression, stress and anxiety [108,118,126,127] were reported as perceived barriers for lifestyle changes. On the other hand, fear of recurrence and perceiving that lifestyle change may prevent recurrence [118,126,128,135,148] was mentioned as a facilitator of lifestyle changes.
One quantitative study found benefit finding to be associated with a significant increase in lifestyle behavior (sleep, diet, physical activity, and stress management), but not with substance use (alcohol consumption and smoking) [79]. Another quantitative study examining optimism found it to be a significant predictor of positive lifestyle behavior change (diet and physical activity) [62].
In the qualitative studies, after treatment side effects [106][107][108]116,118,[125][126][127]135], perceiving no need for lifestyle change [106,108,116], beliefs about (the cause of) cancer being unrelated to lifestyle [106,116,126], low self-efficacy [116,128,134], not enjoying healthy behaviors [107,108,125], and uncertainty about benefits of lifestyle in relation to cancer and health or not perceiving any benefits of lifestyle change [106,116,126,134] were most often mentioned as intra-individual barriers to lifestyle change. On the other hand, in the qualitative studies, the following factors were most frequently mentioned as facilitators at the intraindividual level: cancer diagnosis as wake up call, as initial motivating factor [106][107][108]117,126], perceived/anticipated benefits of lifestyle change (e.g., to improve health, wellbeing, reduce symptoms, improving treatment efficacy & cancer prognosis) [106][107][108]116,118,148], experienced benefits from healthy behaviors (e.g., improved mental wellbeing; help process negative thoughts and feelings) [106,108,116,128,135,148], personal/internal motivation and commitment [107,117,134,135], and goal setting/action planning [108,117,118,125,128]. See Table 3 for an overview of all barriers and facilitators of lifestyle change retrieved from the qualitative studies.

Changes in Sun Protection Behavior
Two quantitative studies reported on determinants of changes in sun protection behavior.

Socio-Demographic Determinants
One of the two studies assessed socio-demographic determinants of changes in sun protection in cancer survivors [78]. Being older than 55 years was found to be significantly associated with increased sun protection behavior as compared to being younger than 55 years [78]. Marital status, employment status, and educational level were not found to be significant predictors of changes in sun protection behavior [78].

Inter-Individual Determinants
None of the studies examined inter-individual determinants of changes in sun protection behavior.

Intra-Individual Determinants
One study assessed intra-individual determinants of changes in sun protection behavior [31]. Fear of cancer recurrence and emotional distress were both not found to be significantly associated with changes in sun protection behavior [31].

Discussion
This systematic review of the literature on psychosocial determinants of lifestyle changes in cancer survivors provides a broad and structured overview of psychosocial determinants per lifestyle behavior on the socio-demographic, inter-individual, and intraindividual level retrieved from both quantitative and qualitative research. To our knowledge, this is the first review on psychosocial determinants of lifestyle change in cancer survivors including qualitative research.
Of the quantitative studies assessing sociodemographic determinants, most assessed educational level as potential determinant of lifestyle change in cancer survivors. These studies mostly showed no association between educational level and lifestyle change [28,33,44,52,59,60,62,70,76,[78][79][80]87,91,[93][94][95]. The studies that did find a statistically significant association showed that higher educational level was associated with more favorable lifestyle changes [32,44,61,84,87,92]. Age and marital status were the next most frequently assessed socio-demographic determinants of lifestyle change in the quantitative studies. These studies showed that marital status was not associated [28,52,53,62,[78][79][80]85,87,91,93,95], as did most of the studies assessing age [28,32,33,44,53,55,[59][60][61][62]64,[78][79][80]85,87]. Ten out of the 35 studies (i.e., 28.6%) assessing age did find an association between age and lifestyle changes. For example, the studies that did find an association between age and smoking behavior suggested older age to be associated with favorable changes in smoking behavior [52,56,95]. In the qualitative studies, ageing was reported as a barrier to being more physically active [98,99,[130][131][132]. Besides ageing, different determinants of lifestyle changes at the sociodemographic level were mentioned in the qualitative studies. Overall, sociodemographic factors were more frequently mentioned as barriers than as facilitators in the qualitative studies. Most qualitative studies mentioned poor weather conditions as a barrier to being more physically active [97][98][99]109,[112][113][114]119,129,131,146,147]. Also, financial constraints (e.g., healthy products being more expensive, costs of using exercise facilities) were mentioned as a barrier to making favorable lifestyle changes [100,106,108,109,118,119,123,132,138,146], while affordability of making lifestyle changes or the financial benefit of smoking cessation was mentioned as a facilitator [100,105,109,140]. Environmental factors (e.g., geographical isolation, lack of footpaths) [108,114,129,132] and work-related factors (e.g., working full-time) [99,113,123,133,145] were also mentioned as socio-demographic determinants in the qualitative studies, primarily as barriers. Overall, our results are in line with and build upon the relatively few studies examining socio-demographic determinants in the review of Park & Gaffey (2007) [24]. As in our study, Park & Gaffey (2007) found that marital status was not associated with lifestyle changes and that the relationship with age and educational level was inconsistent. A systematic review by Kampshoff et al. (2014) [149] examining determinants of physical activity maintenance in cancer survivors, found similar results with no association with marital status, and inconsistent results regarding age and educational level.
Of the quantitative studies assessing determinants of lifestyle changes in cancer survivors at the intra-individual level, self-efficacy was by far the most studied. In those studies, some form of self-efficacy (self-efficacy [27,30,[33][34][35][37][38][39][40]42,45,47,48,50,53,55,58,59,68,73,76,77,86,88,94,95], task self-efficacy [42,43], barriers self-efficacy [27,38,43,74,75], relapse self-efficacy [75], and maintenance self-efficacy [77,82]) was assessed. The vast majority of these studies, assessed associations between self-efficacy and changes in physical activity. More than half of these studies found an association between higher levels of self-efficacy and favorable lifestyle changes. In the qualitative studies, self-efficacy [101,111,116,129,132,135,141,144] was mentioned as a facilitator while low self-efficacy was mentioned as a barrier [100,111,116,119,128,134], primarily of changes in physical activity. Similarly, half of the studies included by Kampshoff et al. (2014) [149] found a positive association between self-efficacy and maintenance of physical activity in cancer survivors, whereas the other half of the studies found no significant associations. In the review by Park & Gaffey (2007) [24] less studies investigated self-efficacy as a potential determinant of lifestyle changes and these studies found mixed results across lifestyle behaviors.
The third most studied potential determinants of lifestyle changes at the intra-individual level in the quantitative studies were anxiety [50,55,60,63,72,87,95] and stages of change [50,[53][54][55]58,76]. Of the quantitative studies that assessed anxiety, four out of seven did not find a statistically significant association with lifestyle changes [50,55,87,95]. The other three studies found an inverse relationship between symptoms of anxiety and favorable lifestyle changes [60,63,72]. Similar results were reported by Park & Gaffey (2027) [24]. Kampshoff et al. (2014) [149] also found anxiety not to be related to maintenance of physical activity. In the qualitative studies, anxiety was also mentioned as a barrier to lifestyle changes as part of the psychological complaints cancer survivors experienced after diagnosis, but also as anxiety specifically related to exercising [99,100,109,137,145,147] and fears related to symptoms [97,98,107,134], while fear of recurrence and the perception that lifestyle change may prevent recurrence was mentioned as a facilitator of lifestyle changes [102,103,113,118,121,123,126,128,135,139,148].
Of the six quantitative studies that assessed stages of change, four found a statistical significant association between a higher stage of change and favorable lifestyle changes, mostly in smoking behavior [53,55,58,76], and one found a borderline significant association [54].
The two studies examining stage of change in the review by Park & Gaffey (2007) [24] found higher stage of change to be related to continued abstinence of smoking and increased physical activity. Stage of change was not mentioned in the qualitative studies.
A frequently mentioned barrier to making lifestyle changes in the qualitative studies was lack of motivation (n = 17) [98,99,101,109,112,114,119,[129][130][131][132]134,135,138,[145][146][147]. Contrary, personal, internal motivation and commitment was mentioned as a facilitator of lifestyle changes [101,107,113,117,122,134,135,137,139,141,143]. In addition, perceiving the cancer diagnosis as a wake-up call or initial motivating factor was mentioned as a facilitator for lifestyle changes in qualitative studies [102,[106][107][108]117,122,126,133]. Motivation was assessed as determinant of lifestyle changes in four quantitative studies, only for changes in physical activity [74,75]. Findings of these studies were inconsistent, with two studies suggesting a positive association between motivation and favorable changes in physical activity [65,69]. In the review by Park & Gaffey (2007) [24], only one study examining motivation was reported, which findings showed an inverse relation to smoking and alcohol consumption.
Some of the included studies examined psychosocial determinants from a theoretical perspective (see Tables S1-S10). For example, some studies studied multiple determinants from Social Cognitive Theory [27,42,62,69,132,150]. This is in line with a previous systematic review showing that lifestyle interventions for cancer survivors have frequently been based on Social Cognitive Theory [151].

Strengths & Limitations
A strength of this systematic review of the literature is the inclusion of both quantitative and qualitative studies. Including both types of research combines the strengths of both research methods and increases the reliability and credibility of the findings [25]. The results demonstrate the added value of including both types of research, clearly showing the differences and similarities in findings from quantitative vs. qualitative research. For example, numerous additional determinants were retrieved from the qualitative studies in addition to the determinants retrieved from the quantitative studies. These additional determinants obtained from qualitative research reflect the cancer survivors' perspective (vs. the predominant researcher's perspective in quantitative studies), which provides additional guidance on how to impact clinical practice and inspires future research.
Another strength is the systematic thorough approach that was applied in this review of the literature. The systematic ordering of the literature per lifestyle behavior provided a detailed overview of the current literature allowing for a specific direction to implications for research and practice. For example, it allows for providing recommendations regarding specific lifestyle behaviors. As each lifestyle behavior is unique, it requires a different health promotion approach. This is illustrated by the observed differences in determinants between lifestyle behaviors.
While interpreting the findings of this review, some limitations should be taken into consideration. Due to the variety in study design of the included studies, we did not conduct a quality assessment. We recommend the reader to incorporate the study characteristics (shown in Tables S1-S10) in interpreting the scientific evidence presented in our systematic review. For example, a large proportion of the included quantitative studies has a cross-sectional study design, whereas either a longitudinal study design or a randomized controlled trial would be preferrable to assess psychosocial determinants of lifestyle changes. In addition, most studies assess lifestyle changes with self-reported data, which could be prone to bias.
This systematic review of the literature provides a wide range of psychosocial determinants of lifestyle change in cancer survivors that can be used to select behavior change techniques and strategies that may be effective in promoting lifestyle change in individual cancer survivors. By matching specific modifiable determinants relevant for this specific patient population to behavior change techniques and strategies, a 'toolbox' containing a variety of building blocks (i.e., intervention ingredients) can be created. The Behavior Change Technique Taxonomy [152], the Behavior Change Wheel [153], and Intervention Mapping [154] could be used to translate these psychosocial determinants into personalized interventions. The importance of such personalized interventions (i.e., personalized lifestyle medicine) is widely recognized nowadays [155,156]. Besides psychosocial factors, many other factors (such as environmental factors on the practice and policy level) may influence lifestyle changes after the diagnosis cancer. Although these factors were not within the scope of this systematic review, they do need to be taken into consideration while promoting lifestyle changes after a cancer diagnosis.
While translating these psychosocial determinants into personalized interventions, the definitions of these determinants should be carefully taken into consideration as differences in definitions may lead to different operationalizations in interventions. In the different included studies, as well as in different theories and models of health behavior change, different terminology may be used to describe similar concepts, such as perceived behavioral control (e.g., defined as "the extent to which a person feels able to perform the behavior" in the Theory of Planned Behavior), perceived competence (e.g., defined as "Seek to control the outcome and experience mastery" in Self-Determination Theory), and self-efficacy (e.g., incorporated in Social Cognitive Theory and in the i-change model) [157]. In some cases, similar terminology is used to describe comparable concepts. For example, self-efficacy is defined as "people's judgements of their ability to cope effectively in different circumstances" according to the Social Cognitive Theory [157], while according to the i-change model, self-efficacy is defined as "a person's perception of their ability to carry out the behavior" [157].
The (oncology) health care provider could play an important role in identifying the (most important) determinants of lifestyle changes in an individual cancer survivor. But, first and foremost, the qualitative results of this systematic review illustrate the important role that oncology health care providers (e.g., oncologists, surgeons) play in changing lifestyle from the cancer survivors' perspective. Our qualitative findings showed that lack of information or advice from health care professionals and lack of knowledge on health benefits were frequently mentioned as barriers to lifestyle changes and that perceived/anticipated benefits were frequently mentioned as a facilitator in the qualitative studies. Oncology health care providers can promote lifestyle changes in the areas in which this is advisable for an individual cancer survivor, by providing evidence-based information and advice on the health benefits of lifestyle change. A source health care professionals could use to obtain evidence-based information about the relation between nutrition, physical activity, and body weight and for evidence-based lifestyle and body weight recommendations for cancer survivors is the website of the World Cancer Research Fund (www.wcrf.org, accessed on 31 January 2022). For cancer survivors, it is important that this information and advice is provided by their oncology health care providers, who they trust and perceive as a credible source, which is a behavior change technique in itself [152]. Other behavior change techniques that could be used to influence some of the determinants that were found to be one of the most influential in this review, include promoting social support by asking cancer survivors about their opportunities for social support in their direct social environment (e.g., social support they could receive from their partner, family or friends) and by advising on, arranging or providing social support (e.g., advise to find a buddy to exercise with) [152]. In addition, self-efficacy could be increased by applying the behavior change techniques goal setting, action planning, graded tasks, (self) monitoring of behavior, and feedback on behavior [152]. Most of these behavior change techniques ((self-)monitoring and feedback on behavior, goal setting, and action planning) were also mentioned as facilitators in the included qualitative studies. These behavior change techniques can be applied by health care professionals during individual counseling sessions which may be supported by digital technology (such as health apps for mobile phones). The use of digital technology may provide a promising means to assist in initiating and maintaining health behavior changes.
While searching for relevant literature for our review, we noticed that we excluded a large amount of quantitative studies (predominantly randomized controlled intervention studies) that did collect the data to be able to study psychosocial determinants of lifestyle changes in cancer survivors, but did not conduct the appropriate analyses to report on determinants of lifestyle changes as this generally was not the primary purpose of these studies. Similarly, we noticed that numerous included quantitative studies typically reported on psychosocial determinants of lifestyle changes using secondary data analyses. In order to further build the evidence base, we recommend to publish such secondary data-analyses in intervention studies that have already collected data on psychosocial determinants. For future intervention studies, it is recommended to, in addition to an effect evaluation, also conduct a process evaluation to gain more insight into (in)effective components of the intervention and mechanisms of behavioral change, and to include psychosocial determinants in data collection and analyses. In addition, large longitudinal observational studies assessing determinants of lifestyle change are valuable means to further build the scientific evidence base. Given the limited amount of included studies on alcohol (n = 4 quantitative; n = 0 qualitative), sun protection (n = 2 quantitative; n = 0 qualitative), and smoking (n = 13 quantitative; n= 4 qualitative) and given the evidence for the positive health effects of making favorable changes in these lifestyle behaviors [11][12][13][14][15][16], future research on psychosocial determinants of these specific lifestyle behaviors is warranted. As almost all of the included studies were either quantitative or qualitative in nature, it would be a valuable addition to conduct more mixed-methods research in this area. Moreover, it would be a valuable addition to conduct studies on psychosocial determinants of lifestyle changes in cancer survivors, using novel techniques, such as Ecological Momentary Assessment, which has the potential of real-life assessment of determinants of lifestyle change.

Conclusions
This overview of the scientific literature on psychosocial determinants of lifestyle change in cancer survivors showed that a large variety of determinants may influence lifestyle change after cancer diagnosis. For example, at the inter-individual level, a positive association between social support and favorable lifestyle changes was found, particularly for changes in physical activity. In addition, advice or support from health care professionals and receiving professional supervision were mentioned as facilitators of favorable lifestyle changes, whereas lack of information or advice from health care professionals was mentioned as a barrier. Psychosocial determinants at the intra-individual level included self-efficacy, psychological complaints (e.g., depression, anxiety, and stress), (lack of) motivation, experienced benefits from healthy lifestyle behaviors, perceived or anticipated benefits of lifestyle change, and receiving knowledge about lifestyle and the effects on health. Findings from this systematic review of the literature demonstrate the important role of oncology healthcare professionals in promoting healthy lifestyle changes in cancer survivors. In addition, findings inform researchers involved in the development of health promotion programs about the methods and strategies they can use to promote healthy lifestyle changes in cancer survivors. Promoting lifestyle change among cancer survivors is expected to have beneficial effects on cancer risk and overall health.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/cancers14082026/s1, Table S1. Overview of quantitative studies assessing psychosocial determinants of changes in exercise/physical activity among cancer survivors (n = 45); Table S2. Overview of quantitative studies assessing psychosocial determinants of changes in diet among cancer survivors (n = 21); Table S3. Overview of quantitative studies assessing psychosocial determinants of changes in smoking cessation among cancer survivors (n = 12); Table S4. Overview of quantitative studies assessing psychosocial determinants of changes in alcohol consumption among cancer survivors (n = 4); Table S5. Overview of quantitative studies assessing psychosocial determinants of changes in lifestyle score among cancer survivors (n = 4); Table S6. Overview of quantitative studies assessing psychosocial determinants of changes in sun protection behavior among cancer survivors (n = 2); Table S7. Overview included qualitative studies on psychosocial determinants of exercise/physical activity only (n = 26); Table S8. Overview of qualitative studies on psychosocial determinants of changes in diet only (n = 9); Table S9. Overview of qualitative papers on psychosocial determinants of changes in smoking only (n = 4); Table S10. Overview of qualitative papers on psychosocial determinants of changes in multiple health behaviors (n = 13).

Conflicts of Interest:
The authors declare no conflict of interest.