A Longitudinal Examination Of The Interrelationships Between Multiple Health Behaviors In Cancer Patients

Paquito BERNARD, Université Laval Cancer Research Center, Québec, Québec, Canada, Canada

1 Université Laval Cancer Research Center, Québec, Québec, Canada


Cancer survivors are at a greater risk of developing secondary cancers and other chronic diseases. Health behaviors (HBs) are determinant protective factors of cancer recurrence and against other chronic illnesses. Understanding the possible interrelationships between HBs during and after cancer treatment is crucial to orient the development of more effective interventions to change HBs that are more effective.

The present study assessed the relationships between self-reported smoking, physical activity, alcohol intake and caffeine consumption among cancer patients with mixed cancer sites followed for 18 months. We explored the temporal associations of each HB with all three other HBs.


Patients completed an HB questionnaire at the perioperative period and then 2,6,10,14,18 months after. The interrelationships between HBs were examined using cross-lagged analyses conducted with structural equation modeling. The model was adjusted for age, sex, income, and cancer diagnosis, and for time-varying variables: weight, chemotherapy and radiotherapy.


The study included 962 participants. The most frequent cancer sites were breast (49.4%), prostate (28.3%) and gynaecologic (10.8%). Although the model showed a good fit to the data, SRMR=0.050, RMSEA=0.060, and CFI=0.85, no significant cross-temporal paths emerged between HBs. However, higher levels of physical activity at 14 and 18 months were significantly predicted by a lower nicotine and alcohol consumption, and a greater caffeine intake at 14 and 18 months. For all four HBs, continuity paths generally indicated that one particular HB was significantly predicted by the same HB at the previous time point.


This study revealed that HBs assessed following cancer surgery are mostly independent. These findings suggest that interventions promoting HB changes during the cancer treatment trajectory could target several HBs simultaneously without any interference between them. They also suggest that physical activity is influenced by other HBs but only after when the cancer treatment is completed.