Oncology · Medical Literature
Diet and Physical Activity as Mediators of Weight Loss in a Digital Prostate Cancer Survivorship Intervention for Men: Secondary Analysis of the PC-PEP Randomized Trial.
Abstract
Weight gain during prostate cancer treatment can worsen treatment-related side effects, impair quality of life, and increase risks of recurrence and prostate cancer-specific mortality. The Prostate Cancer Patient Empowerment Program (PC-PEP), a comprehensive, home-based 6-month digital intervention, has demonstrated improvements in diet, physical activity, and weight. This secondary analysis examined whether changes in diet and physical activity mediated weight loss among participants receiving PC-PEP. In a randomized clinical trial, 128 men with localized prostate cancer were assigned to PC-PEP (n=66) or standard care (n=62) for six months. Diet quality was assessed using a 12-item summary score reflecting weekly frequency of key food groups. Physical activity level was self-reported as not very active, moderately active, or very active. Mediation models were adjusted for age, comorbidity, treatment modality, time from randomization to treatment, relationship status, medication for anxiety or depression, and baseline weight, diet, and physical activity. The results show that weight loss in the PC-PEP group was significantly mediated by both diet and physical activity, accounting for 21% and 16% of the total effect, respectively. The direct effect of the intervention remained significant. These results indicate that improvements in diet and physical activity partially mediated weight loss in men receiving PC-PEP, while a substantial direct effect suggests that additional components of the intervention contribute to outcomes. Findings highlight the value of comprehensive, multi-faceted digital survivorship interventions for men with prostate cancer and suggest that observed mediation effects are conservative estimates given the use of brief self-reported behavioral measures.
Clinical relevance
Do improvements in diet quality and physical activity level mediate the weight loss observed in men with localized prostate cancer who receive the PC-PEP digital survivorship intervention compared with standard care? __SNAPSHOT__ {"studyDesign":"Secondary mediation analysis of a randomized controlled trial (crossover/waitlist-controlled design); not a primary efficacy trial","setting":"Single institution; Nova Scotia, Canada (Halifax); participants recruited December 2019 – January 2021 via urologists, radiation oncologists, and clinic advertisements","sampleSize":"n=128 (PC-PEP: n=66; waitlist-control/standard care: n=62)","population":"Men aged 50–82 years with biopsy-proven localized prostate cancer (adenocarcinoma) scheduled for curative-intent surgery or radiotherapy (±hormone therapy); recruited across major oncology clinics in Nova Scotia","intervention":"PC-PEP: a 6-month home-based digital program comprising 182 daily emails with video-based exercise (30-minute strength training using body weight and elastic bands, aerobic activity, pelvic floor exercises) and predominantly plant-based dietary education, live online events, stress reduction via biofeedback, and social/emotional support","comparator":"Standard of care for 6 months (waitlist-control; late-crossover group subsequently received PC-PEP at months 6–12)","primaryOutcome":"Weight change (kg) at 6 months, with diet quality (12-item summary score reflecting weekly frequency of key food groups) and self-reported physical activity level (3-category ordinal: not very active, moderately active, very active) tested as mediators","followUp":"6 months (primary mediation window); 12-month data available for the parent trial","fundingSource":"Not reported in the supplied abstract; prior PC-PEP publications acknowledge institutional and government sources — confirm in full text"}
Key findings
The abstract states that weight loss in the PC-PEP group was 'significantly mediated' by both diet quality and physical activity level; however, the abstract is truncated and specific effect sizes, indirect effect estimates, proportion of total effect mediated, confidence intervals, bootstrapped confidence intervals, and p-values for the mediation paths are not reported in the supplied text. From the parent trial (prior secondary analysis, Current Oncology 2024), the PC-PEP early intervention group lost a mean of 2.7 kg over 6 months versus waitlist control (p<0.001). The present mediation analysis adjusts for age, comorbidity (Charlson Comorbidity Index), treatment modality, time from randomization to treatment start, relationship status, medication for anxiety or depression, and baseline weight, diet, and physical activity. Diet was assessed using a 12-item summary score; physical activity was self-reported on a 3-point ordinal scale. Quantitative mediation effect sizes, indirect effect confidence intervals, and proportion of variance explained are not reported (abstract truncated). No adverse events were reported in the parent trial. NNT and NNH are not calculable from the available data.
Practice implications
This secondary mediation analysis provides mechanistic insight into how PC-PEP achieves weight loss, suggesting that the intervention's effects on weight are not direct but operate through improvements in dietary behavior and physical activity. This is clinically meaningful because it identifies the behavioral pathways that should be targeted and preserved in any adaptation or scale-up of the program. However, several important limitations must be recognized before practice implications are drawn. First, only an abstract (truncated) was available for review; full mediation coefficients, confidence intervals, and model diagnostics cannot be independently verified. Second, this is a secondary analysis of a relatively small single-center RCT (n=128), underpowered for the complex mediation modeling performed. Third, physical activity was measured using a coarse 3-category self-report scale, which lacks the sensitivity to capture dose-response relationships or distinguish aerobic from resistance activity. Fourth, diet quality was assessed via a non-validated 12-item summary score rather than a validated dietary assessment instrument (e.g., 24-hour recalls, validated food frequency questionnaire), introducing measurement error in the proposed mediator. Fifth, both mediators are self-reported and subject to social desirability bias in a digital intervention context. Sixth, the sample was drawn from a single Canadian province and was predominantly composed of men undergoing curative treatment, limiting generalizability to men on active surveillance, those with advanced disease, or those from diverse racial/ethnic backgrounds. Seventh, the crossover design means some contamination of the control arm is possible, potentially attenuating between-group differences. The findings do not establish causality between the mediators and long-term oncologic outcomes such as recurrence, cancer-specific survival, or overall survival. Clinicians should interpret these findings as hypothesis-generating and mechanistically supportive, not as definitive evidence that dietary or physical activity improvements independently cause the observed weight loss in a manner that will translate to survival benefit. __LIMITATIONS__ Abstract-only review: quantitative mediation parameters, indirect effect sizes, and confidence intervals are not available for critical appraisal,Secondary mediation analysis of an RCT not originally powered for mediation — risk of type II error for individual paths,Small single-center sample (n=128) from a single Canadian province; limited generalizability across ethnically diverse or geographically varied populations,Physical activity assessed via coarse 3-level self-report ordinal scale, not objective accelerometry or validated questionnaire,Diet quality assessed via a non-standard 12-item summary score; validity and psychometric properties not described in the abstract,Both mediators are self-reported and susceptible to recall bias and social desirability in a digital intervention,Waitlist-crossover design: late-group crossover may introduce contamination or dilute control-arm contrasts,Mediation models assume temporal precedence of mediator change before weight change, which cannot be confirmed from a 6-month assessment without intermediate time points,Absence of objective weight verification beyond online survey self-report (prior analyses note substantial missing in-person measurement data),No long-term oncologic outcomes (recurrence, cancer-specific mortality) assessed; weight loss as a surrogate endpoint for survival benefit is unproven in this context,Funding source not confirmed in abstract; potential for sponsor influence cannot be assessed __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ This secondary mediation analysis of the PC-PEP RCT suggests that a 6-month digital prostate cancer survivorship program achieves clinically meaningful weight loss (~2.7 kg versus standard care) through improvements in both diet quality and physical activity, identifying these as actionable behavioral targets. However, because the abstract is truncated, mediation effect sizes cannot be fully appraised, and the small single-center sample limits generalizability. Confirmation in larger, multicenter trials with objective physical activity measurement, validated dietary instruments, and long-term oncologic endpoints is needed before routine clinical integration.