Further evidence for the co‐occurrence of adverse health behaviors among childhood cancer survivors

To the Editor: We read with interest the report from the Childhood Cancer Survivor Study (CCSS) by Phillips-Salimi et al. [1] examining health behaviors in childhood cancer survivors. A pattern of several risky health behaviors was evident in their data, with survivors being more likely to currently smoke and engage in HIV-related risk-behaviors and less likely to engage in leisure-time physical activity compared to healthy controls. We recently analyzed retrospective data on health behaviors for 208 adult survivors of childhood cancer treated at Sydney Children’s Hospital, Australia, up to 30 years post-diagnosis [2,3]. Compared to the CCSS sample, our sample was slightly younger (median age: 27.9 years, range: 18–48), comprised mainly of leukemia, Hodgkin, or non-Hodgkin lymphoma survivors (60.1% combined, compared with only 10.8% in the CCSS), and had a greater proportion of male respondents (37.5% vs. 18.7%). In our study, survivors completed questionnaires assessing several health-behavioral indicators, including cigarette smoking, alcohol consumption, and BMI. Survivors’ reported alcohol consumption was compared to the Australian National Health and Medical Research Council guidelines for drinking that may lead to short-term harm (consuming 5–6 drinks/day for females and 7– 10 drinks/day for males) [4], while our definition of an Obese BMI was identical to the CCSS (BMI >30 kg/m). Several of our findings echo the report by Phillips-Salimi et al. A minority (11%) reported risky levels of alcohol consumption [4], and fewer survivors reported current smoking than in previous CCSS reports (13% vs. 17%) [5], and the general population (19% daily smokers in 2007–2008) [6]. Half of our sample had a normal-range BMI (53.7%). Adverse health behaviors also appeared to co-occur in our sample: survivors who smoked were 5.3 times more likely to engage in risky drinking than survivors who did not smoke (OR 1⁄4 5.3; 95% CI 1⁄4 1.8–15.1, P 1⁄4 0.002). The late effects that survivors develop are largely dependent on the cancer therapies received during childhood. In our analyses, no differences in smoking habits or alcohol consumption were evident among survivors who received different treatments, despite being in vulnerable subgroups. Survivors treated with pulmonary and/or cardiotoxic agents were less likely to be overweight/obese (OR 1⁄4 2.5, 95% CI 1⁄4 1.3, 4.8, P 1⁄4 0.007), but neither their smoking (OR 1⁄4 1.209, 95% CI 1⁄4 0.8–1.8, P 1⁄4 0.415) nor drinking (OR 1⁄4 1.134, 95% CI 1⁄4 0.5–2.7, P 1⁄4 0.814) differed significantly from their counterparts. The cardiotoxic effects of anthracyclines may increase over time in cancer survivors, with the cumulative incidence of clinical heart failure reaching 5% after 15 years [7]. Consequently, smoking may be of particular significance to survivors who are at high risk of cardiac dysfunction [8] and at risk for second malignancies [3]. The finding that smoking and ‘‘risky’’ drinking behaviors were unaltered by treatment regimen suggests that these behaviors are primarily driven by the same attitudes observed in the general population. Consideration needs to be given to the efficacy of educational strategies to alter socially acceptable behaviors in cancer survivors. Given that the CCSS survivors indicated poorer perceived social/emotional support [1], addressing the psychosocial context in which survivors are expected to enact positive health behaviors may be crucial to support long-term change.