Post-diagnostic health behavior scores in relation to fatal prostate cancer

Published in Cancer

What We Knew

Prostate cancer (PCa) is the second most common cancer among men worldwide, with nearly 1.5 million cases diagnosed each year.1 Though most men are diagnosed with local or regional disease,2 PCa becomes fatal for roughly 375,000 men annually.1 As such, a man diagnosed with PCa is likely to ask the question, “What can I do to reduce my risk of death from my disease?” Given that a cancer diagnosis is a “teachable moment”3-5 – a period during which an individual may be open to behavior change – providers should be equipped with evidence-based, disease-specific lifestyle recommendations.

Individual post-diagnostic health behaviors, such as smoking, exercise, and diet, have been associated with the risk of fatal PCa. In addition, scores comprised of pre-diagnostic health behaviors in combination have been associated with the risk of aggressive PCa.6-9

What We Didn’t Know

But what about post-diagnostic behaviors in combination? No prior studies had examined health behavior scores after diagnosis in relation to PCa death. We thus set out to evaluate four health behavior scores in the post-diagnostic context: 1) a novel, literature-based, three-factor score comprised of behaviors that have been associated with fatal outcomes (“2021 PCa Behavior Score”); 2) a six-factor score developed for the prevention of lethal PCa (“2015 PCa Behavior Score”);6 3) a score developed based on World Cancer Research Fund (WCRF)/​American Institute for Cancer Research (AICR) diet and exercise guidelines for cancer prevention (“WCRF/AICR Score”);10 and 4) a score developed based on American Cancer Society diet and exercise recommendations for cancer survivorship (“ACS Score”).11 

What We Did

We used men diagnosed with non-metastatic PCa from the Health Professionals Follow-up Study to conduct a prospective cohort study of post-diagnostic health behavior scores and risk of fatal disease. Participants self-reported smoking status, body mass index, and physical activity every two years during follow-up. They reported on diet via food frequency questionnaires administered every four years. The health behaviors contributing to each score are summarized by the following table:


2021 PCa Behavior Score

2015 PCa Behavior Score


ACS Score

Smoking Status



Body Mass Index

Physical Activity





Red and Processed Meat


Whole Grains




Fruits and Vegetables






Fatty Fish




Sugar-Sweetened Beverages




Ultra-Processed Foods





Higher values of the scores corresponded to greater adherence to recommended behaviors. We used multivariable Cox proportional hazards regression models to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for each post-diagnostic score and fatal PCa. In our primary analyses, we lagged exposure four to six years relative to each period of follow-up. In secondary unlagged analyses, exposure corresponded to the most recent questionnaire for the follow-up period. The latter analyses considered more acute impacts of health behaviors on PCa outcomes but were more susceptible to reverse causation.

What We Found

Primary analyses of the 2021 PCa Behavior Score were based on 4362 men diagnosed with stage T3a or lower PCa, followed for a median (interquartile range) of 10 (6.0, 15) years. There were 219 deaths from PCa. The absolute crude rates of fatal PCa per 1000 person-years were lower for those with 3 points (4.1) than 0-1 points (6.7). However, the 2021 PCa Behavior Score was not associated with fatal PCa in models adjusted for clinical characteristics and pre-diagnostic behaviors (HR3 pts vs. 0-1 pts: 0.84, 95% CI: 0.46-1.52, Ptrend: 0.77).

Primary analyses of the 2015 PCa Behavior Score were based on 4320 men, a median follow-up of 8.2 (5.3, 13) years, and 183 fatal events. The absolute crude rates of fatal PCa per 1000 person-years were 3.9 versus 6.5 for those with 5-6 points versus 0-1 points, respectively. The score was trended significantly inversely associated with the risk of fatal PCa (HR5-6 pts vs. 0-1 pts: 0.68, 95% CI: 0.29-1.57, Ptrend: 0.02), though event numbers were limited in extreme categories of the score. Each one-point increase in the score was associated with a 19% reduced risk. 

Primary analyses of the WCRF/AICR Score were based on 4518 men and 208 fatal events, and those for the ACS Score were based on 4505 men and 207 fatal events. Both analyses included a median follow-up of 8.7 (5.7, 14) years. Neither score demonstrated evidence of an association with fatal PCa.

Sensitivity analyses based on simple updating (rather than lagged) exposures yielded statistically significant inverse associations for the 2021 PCa Behavior Score (HR3 pts vs. 0-1 pts: 0.41, 95% CI: 0.26-0.65, Ptrend<0.001), 2015 PCa Behavior Score (HR5-6 pts vs. 0-1 pts: 0.68, 95% CI: 0.34-1.36, Ptrend: 0.01), and ACS Score (HR5-6 pts vs. 0-1.5 pts: 0.57, 95% CI: 0.34-0.96, Ptrend: 0.02), and a suggestive inverse association for the WCRF/AICR Score (HR5.25-7 pts vs. 0-2.5 pts: 0.71, 95% CI: 0.44-1.15, Ptrend: 0.13).

 What We Learned

In this population of men diagnosed with nonmetastatic PCa, each one-point increase toward not smoking, maintaining a healthy body size, engaging in regular physical activity, and consuming specific food items (2015 PCa Behavior Score) was associated with a 19% lower risk of dying from PCa. Other scores did not demonstrate convincing evidence of associations with fatal PCa.

 The sum of our results suggests that: 1) dietary factors may contribute less to risk reduction than the combined effects of not smoking, exercise, and body size; and 2) there may be very specific dietary factors or patterns associated with PCa mortality. While it remains possible that aspects of diet not studied here could be important, the overall dietary patterns captured by the WCRF/AICR and ACS Scores were not strongly associated with PCa outcomes in this population. That said, a diet that helps men maintain a healthy body weight and physical activity seems to be important.

 Timing of healthy behaviors also seems to be important. Whereas analyses relating the 2021 and 2015 PCa Behavior Scores to fatal events in the same two-year period yielded strongly inverse signals, lagged analyses of four to six years produced attenuated results. Were worsening PCa to result in less healthy behaviors, the result would be an inverse relationship between the scores and fatal PCa. That the lagged analyses were attenuated could suggest such reverse causation. Alternatively, it could be that the relevant etiologic period for exposure is close in time to potential outcomes. Then, we would not expect to see a relationship when evaluating an exposure window occurring years prior to a fatal event.

 In summary, for men diagnosed with nonmetastatic PCa, adhering to recommendations to avoid tobacco, maintain a healthy body size, and engage in regular physical activity may decrease the risk of dying from PCa.

 What Is Next

Further research is needed to understand our results in the context of prior associations between individual behavioral risk factors and fatal PCa and to inform tailored cancer survivorship recommendations.


1          Ferlay J, Colombet M, Soerjomataram I, Parkin DM, Piñeros M, Znaor A, et al. Cancer statistics for the year 2020: An overview. Int J Cancer. 2021 10.1002/ijc.33588

2          Siegel RL, Miller KD, Fuchs HE & Jemal A. Cancer statistics, 2022. CA Cancer J Clin. 2022;72:7-33.

3          Bell K. Remaking the self: Trauma, teachable moments, and the biopolitics of cancer survivorship. Cult Med Psychiatry. 2012;36:584-600.

4          Ganz PA. A teachable moment for oncologists: Cancer survivors, 10 million strong and growing! J Clin Oncol. 2005;23:5458-60.

5          Karvinen K, Bruner B & Truant T. The teachable moment after cancer diagnosis: Perceptions from oncology nurses. Oncol Nurs Forum. 2015;42:602-9.

6          Kenfield SA, Batista JL, Jahn JL, Downer MK, Van Blarigan EL, Sesso HD, et al. Development and application of a lifestyle score for prevention of lethal prostate cancer. J Natl Cancer Inst. 2016;108:djv329.

7          Olmedo-Requena R, Lozano-Lorca M, Salcedo-Bellido I, Jiménez-Pacheco A, Vázquez-Alonso F, García-Caballos M, et al. Compliance with the 2018 world cancer research fund/american institute for cancer research cancer prevention recommendations and prostate cancer. Nutrients. 2020;12

8          Romaguera D, Gracia-Lavedan E, Molinuevo A, de Batlle J, Mendez M, Moreno V, et al. Adherence to nutrition-based cancer prevention guidelines and breast, prostate and colorectal cancer risk in the mcc-spain case-control study. Int J Cancer. 2017;141:83-93.

9          Arab L, Su J, Steck SE, Ang A, Fontham ET, Bensen JT, et al. Adherence to world cancer research fund/american institute for cancer research lifestyle recommendations reduces prostate cancer aggressiveness among african and caucasian americans. Nutr Cancer. 2013;65:633-43.

10        World Cancer Research Fund/American Institute for Cancer Research. Diet, nutrition, physical activity and cancer: A global perspective. Continuous update project expert report 2018. (2018).

11        Rock CL, Doyle C, Demark-Wahnefried W, Meyerhardt J, Courneya KS, Schwartz AL, et al. Nutrition and physical activity guidelines for cancer survivors. CA Cancer J Clin. 2012;62:243-74.


Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in

Subscribe to the Topic

Cancer Biology
Life Sciences > Biological Sciences > Cancer Biology

Related Collections

With collections, you can get published faster and increase your visibility.

Digital Imaging

BJC’s Digital Imaging series is open to receiving submissions assessing:
  • State-of-the-art in digital imaging technology and computational output;
  • Radiomics;
  • Algorithms and approaches designed to predict therapeutic response and enable “personalized medicine”;
  • Interrogation of the immune microenvironment and implications for immunotherapy selection;
  • Challenges of implementation in clinical practice;
  • Controversial applications of digital imaging

Publishing Model: Hybrid

Deadline: Ongoing