Effects of high-intensity training on the quality of life of cancer patients and survivors: a systematic review with meta-analysis

Cancer and associated medical treatments affect patients' health-related quality of life (HRQoL) by decreasing functional dimensions of physical, social, cognitive, and emotional well-being, while increasing short and late-term symptoms. Exercise, however, is demonstrated to be a useful therapy to improve cancer patients' and survivors’ HRQoL, yet the effectiveness of high-intensity training (HIT) exercise is uncertain. This systematic review and meta-analysis aimed to analyse the effects of HIT on HRQoL dimensions in cancer patients and survivors as well as evaluate the optimal prescription of HIT. The search followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (PRISMA) and examined Web of Science and PubMed (Medline) databases. Data were analysed utilizing Review Manager Software. Twenty-two articles were included in the systematic review and 17 in the meta-analysis. Results showed HIT improved global quality of life, physical functioning, role functioning, social functioning, cognitive functioning, fatigue, pain, dyspnea, and insomnia, compared to an inactive control group, yet no differences were found between HIT and low to moderate-intensity exercise interventions. Particular improvements in HRQoL were observed during cancer treatment and with a training duration of more than eight weeks, a frequency of 2 days/week, and a volume of at least 120 min/week, including 15 min or more of HIT. Our findings whilst encouraging, highlight the infancy of the extant evidence base for the role of HIT in the HRQoL of cancer patients and survivors.

. To make a representative analysis, the EORTC QLQ-C30 dimensions were used to group FACT and SF-36 results. We examined all the questionnaire's dimensions and related items to establish similarities between the categories. We did this even if there were named differences, but their items evaluated the same topic. Item categorization was analyzed by two of the researchers (AMLP and DCM) who discussed similarities and differences to classify them in useful variables for the meta-analysis. The variables were divided into categories consistent with EORTC QLQ.C30 dimensions, distinguishing global health, functional scale, and symptoms scale. The data from those items that did not correspond to any variable group created, or was registered by less than three articles, was not included in the literature part.
Statistical analysis. Post-intervention means and standard deviations were extracted from the articles and analyzed using Review Manager Software (RevMan, 5.3) 47 based on; High-intensity exercise group (HIEG), low-to-moderate exercise group (LMEG), and inactive control group (CG). When outcomes were evaluated on scales with opposite directions, (e.g. pain or fatigue), one of the results directions was multiplied by − 1 48 . The results were reported using standardized mean differences (SMDs) and interpreted according to the Cochrane Handbook 48 i.e. small effects with scores < 0.4, moderate effects from 0.4 to 0.7, and large effects with > 0.7. The statistical method employed was inverse variance with random effects 49 and the interval confidence (CI) utilized was 95%. Different analyses were computed for each dimension (Global health, Physical functioning, Role functioning/physical role, Emotional functioning/wellbeing and mental health, Cognitive functioning, Social functioning, Fatigue/vitality, Nausea, Body pain, Dyspnoea, Constipation, Insomnia, Diarrhoea and Appetite loss). The described procedure was carried out, first, to analyze the difference according to the type of intervention group: LMEG or CG. Secondary calculations were performed contrasting HIEG and CG outcomes with more detail making the following subgroups analysis: (1) interventions conducted before, during, or after cancer treatment, (2) interventions of ≤ 8 weeks or > 8 weeks, (3) only aerobic exercise programs or work-outs with any resistance component, (4) studies where participants exercised ≤ 2 times per week or those > 3 times per week (5) interventions of ≤ 120 min or > 120 min per week, (6) training designs with a high-intensity aerobic session part of 15 min or less and separately those with greater than 15 min duration.

Results
Study selection. Figure 1 sets out the data from the study selection process. The search obtained 385 articles, 157 in the PubMed database and 228 in Web of Science. Two more papers were identified in the references of articles and were therefore included 50,51 . One hundred thirty-five of the found studies were duplicated, so 251 were screened by examining the title and abstract. Following the exclusion criteria, one animal intervention,  Table 1. Risk of bias using PEDro scale. Y: Yes, the item was satisfied in the experimental protocol; N: No, the item was not satisfied in the experimental protocol Items: (1) Eligibility criteria; (2) Random allocation; (3) Concealed allocation; (4) Similarity of the groups at baseline; (4, 7-11) Key outcomes; (5-7) Blinding process; (8) Final measure with 85% of the initial sample size; (9) intent-to-treat analysis; (10) Between-group comparisons report; (11) Point and variability measures.
Physical exercise and treatment timing. When the HIEG exercise programs were implemented during cancer treatments, physical functioning (p = 0.0005, with SMD of 0.42 and a 95% CI from 0.18 to 0.66) role functioning (p = 0.0003, with SMD of 0.35 and a 95% CI from 0.16 to 0.54) and social functioning (p = 0.03, with SMD of 0.12 and a 95% CI from 0.01 to 0.23) seemed to improve more than the CG. In contrast, those variables were not significantly improved from after-treatment interventions. Moreover, outcomes of global health dimensions showed similar between-group differences in exercise programs conducted during (p = 0.02, with SMD of 0.22 and a 95% CI from 0.03 to 0.40) and after (p = 0.003, with SMD of 0.30 and a 95% CI from 0.10 to 0.50) cancer treatments.
Intervention length. Results showed higher between-group differences (HIEG vs CG) when performing HIT in exercise programs of ≤ 8 weeks duration, including physical function (p = 0.04, with SMD of 0.04 and a 95% CI from 0.01 to 0.45) and role functioning (p = 0.02, with SMD of 0.26 and a 95% CI from 0.04 to 0.49). For HIT programs lasting more than 8 weeks there was no significant between-group differences (physical function p = 0.05, role functioning p = 0.07). However, cognitive functioning reached higher significant differences between the CG and the EG in interventions longer than 8 weeks (p = 0.04, with SMD of 0.20 and a 95% CI from 0.01 to 0.40). All HIT durations showed significant differences between CG and HIEG in the global health dimensions (≤ 8 weeks: p = 0.04; > 8 weeks: p = 0.002) and fatigue (≤ 8 weeks: p = 0.008; > 8 weeks: p = 0.001). www.nature.com/scientificreports/ Exercising frequency. The physical (p = 0.005, with SMD of 0.37 and a 95%CI from 0.11 to 0.62) and cognitive functioning (p = 0.003, with SMD of 0.25 and a 95% CI from 0.08 to 0.42) dimensions showed significant between-group differences with higher improvements in HIEG vs CG in interventions conducted 2 times/week. In comparison, 3 times/week programs did not show significant between-group differences (physical p = 0.09; cognitive p = 0.18). However, patients in the HIEG scored higher than the CG in role functioning with a frequency of 3 times/week (p = 0.04, with SMD of 0.21 and a 95% CI from 0.01 to 0.42), while no significant results were observed in interventions with lower frequency (p = 0.05). All of the reported exercise frequencies showed significant differences between CG and HIEG in global health (2 times/week: p = 0.002; 3 times/week: p = 0.03) and fatigue (2 times/week: p = 0.005; 3 times/week: p = 0.001) dimensions. The meta-analysis did not include the global calculation of HRQoL because of the data heterogeneity from the different questionnaires' measures, despite this there were significant improvements in most of the articles analyzed [54][55][56]59,62,69,74,75 as the Supplementary Tables S4 and S5 report.

Discussion
This study aimed to explore the effect HIT on HRQoL dimensions in cancer patients and survivors. We also aimed to evaluate the optimal characteristics of HIT for dimensions of HRQoL with respect to intervention timing and cancer treatment, mode of exercise, and exercise dose. We found that HIT improves global quality of life, physical functioning, role functioning, social functioning, cognitive functioning, fatigue, pain, dyspnoea, and insomnia, compared to an inactive control group. The inclusion of resistance training seemed critical to improvements in global health and physical functioning. No significant differences were found when the effects of HIT were compared to low to moderate-intensity exercise. Improvements in HRQoL were observed during cancer treatment when training occured for more than eight weeks, with a frequency of 2 days/week, and a volume of at least 120 min/week with the HIT component duration in each session of at least 15 min.
Global health and physical function were the most commonly reported variables studied in exercise and cancer reviews, and findings here suggest that HIT consistently shows improvements in these outcomes compared to an inactive control group 32,33 . Data support positive global health changes with intense exercise 37 , but are contrary to Mugele et al. (2019), who focused solely on High Intensity Interval Training (HIIT) in their systematic review 38 . The broader definition of HIT might explain the data we observed here, but it is clear further studies are required to understand the role of HIT, including HIIT, specifically on HRQoL outcomes in cancer.
The subgroup analysis made regarding an intervention's timing showed statistically positive effects in the global health dimension, physical functioning, role functioning, and social functioning during cancer treatments. Only the global health dimension showed a positive increase in after-treatment HIT. In line with our findings, the functional variables of HRQoL decrease progressively across chemotherapy 78 . Our data suggest that this decline might be moderated with HIT, particularly regarding depression and anxiety, function 79 and activities of daily living 80 . Most of the negative side effects of cancer and its treatments are related to reduced physical functioning, reduced mobility due to surgery or chemotherapy 81 , lymphedema 82 , negative body composition changes as sarcopenia 83 , or osteoporosis 84 . Providing opportunities to mitigate these deleterious effects through HIT is highly important since more than half of all cancer patients develop a mobility disability because of the disease and its treatments' adverse side-effects 85 86 . Exercise has been shown to increase baseline physical functioning, reduce treatment-related impediments 30 , and help a patient maintain overall strength during treatment 87 . Post-treatment exercise can help the patient return to baseline and reduce subsequent side-effects 88 . The meta-analysis underlined the important role of resistance training in improving global health and physical function. Incorporating strength training in HIT programs is likely to increase muscle function, reduce the risk of sarcopenia, and reduce the risk of mortality 89 and treatment toxicity 90 . This has been shown independent of age, cancer stage, or BMI 91 , and is partly explained through an anti-inflammatory response 92 . Further, resistance training may regulate deficiencies in skeletal muscle and adipose tissue known as cachexia 93,94 . However, it should be noted that interventions which included resistance training had lower adherence rates compared to aerobic exercise, which has been reported in other chronic disease patients 95 . Poor adherence might also explain why social and emotional functioning only significantly increased in the aerobic component programs, not in the resistance exercise modalities. To improve adherence, researchers and exercise specialists might wish to adopt a co-production approach, seeking to co-create the specific training strategies with people who have a cancer diagnosis, taking into account what matters most to them 95 .
Cancer and its associated treatments can cause severe side-effects during drug therapy, with pain and fatigue the most common 96 . Fatigue-related to cancer is reported by 70% of cancer patients 97 . The complaint of cancerrelated fatigue is associated with immune response dysregulation, inflammation, metabolic and mitochondrial function impairment, neuroendocrine function impairment, and genetic biomarkers 98 ; however, with exercise, these parameters can be improved 99 . To decrease fatigue, HIT, as well as other exercise modes, seems to be effective 30 , possibly more so than pharmacological or psychological therapies 100 . Other symptoms like pain, insomnia, and dyspnea also appear to improve via exercise 35 and without aggravating cancer symptoms, although this requires further investigation 101 .
Interventions lasting more than 8 weeks reported greater increases in HRQoL compared to shorter duration programs, which is consistent with a previous review of HIT interventions 37 . Greater improvements across a range of cancer-related outcomes were observed with exercising 3 times/week compared to training 2 times/week, except for role functioning (3 times/week). The American College of Sports Medicine recommends exercising two to three times/week 101 , which agrees with the findings of our study and a previous meta-analysis 37 . Three exercise sessions per week will also make it easier for individual cancer patients to achieve 120 min of weekly exercise, which seems to be important for increasing HRQoL, particularly when each session includes at 15 min of HIT. Some programs have included family members with hospitalized patients 102 .
This article presents valuable information about the role of high-intensity exercise as part of treatment and recovery in cancer, specifically in terms of HRQoL. The data from the systematic review and meta-analysis should be viewed in the light of the following limitations. Only articles written in English or Spanish were included, so not all the available information was analyzed. The intervention description, as well as the subgroup meta-analysis, was undertaken with the published available details. Where a study had incomplete data (e.g. sessions' duration, HIT minutes, after intervention mean and standard deviation, etc.), data were omitted to the corresponding subgroup calculation. For the meta-analysis procedure, data from at least three articles were needed to make a subgroup analysis. Thus, assessments concerning the cancer type and all the subgroups analysis considering each intervention characteristic were not possible. Therefore, more information could be added with further studies. It must be considered that three of the included articles combined HIT with body awareness, relaxation, or massage interventions, each of which could influence HRQoL.

Conclusion
This is the first meta-analysis exploring the effects of HIT on the HRQoL of cancer patients and survivors. Data from this systematic review and meta-analysis suggests that HIT as part of exercise therapy for people with a cancer diagnosis can improve global health and provide physical, cognitive, and social functioning benefits compared to controls. In addition, fatigue, bodily pain, dyspnea, and insomnia decreases can be achieved with HIT, all with similar outcomes observed using low-moderate intensity exercise. Dimensions of HRQoL showed the largest positive effects when the programs were delivered as part of cancer treatment and included resistance training. Ultimately, exercise programs may need to be longer than 8 weeks, with a HIT frequency of 2 times/week, and a total duration of at least 120 min/week, including a HIT component of more at least 15 min, to achieve the highest return in HRQoL. However, as it is the first meta-analysis about the effects of HIT in the HRQoL of cancer patients and survivors, further research is required to support our findings.