Potential for risk reduction of chronic health conditions through lifestyle in childhood cancer survivors

Figures

• 

  Fig. 1. The central illustration summarizes the population attributable fractions (PAF) of each chronic health condition attributed to unfavorable lifestyle score among participants in the Childhood Cancer Survivor Study, n = 18,664. Coloring of symbols indicate the individual lifestyle variables most strongly contributing to the results, with lack of exercise in green, overweight/obesity in blue, smoking in red and heavy/risky drinking in yellow. Chronic health conditions not associated with the lifestyle score are not colored and the estimation of PAF is not applicable (NA). The figure also indicates for which chronic health conditions the PAF of lifestyle score exceeds those of chemotherapy (triangle) or radiotherapy (circle). Stars indicate that the absolute excess risk of not having a healthy lifestyle score is significantly higher in childhood cancer survivors than in sibling controls. Created in BioRender. Onerup, A. (2026) https://BioRender.com/jge2wr2.

• 

  Fig. 2. Rate ratios for unhealthy and moderately healthy lifestyle scores, in reference to healthy lifestyle score, for chronic health conditions. Bars represent rate ratios with 95% confidence intervals. Piecewise exponential models were used. All analyses were adjusted for attained age as cubic splines, sex, race/ethnicity, age at cancer diagnosis, socioeconomic status (including education, marriage status, household income, and insurance), and treatment variables chosen from backward selection for each outcome. All statistical tests were two-sided. No correction was performed for multiple testing since all outcomes were distinct hypotheses of individual interest. Number of participants in the total population is 18,664, with analysis of each outcome differing slightly due to pre-existing conditions, as reported in Supplementary Table 1. Source data are provided as a Source Data file. CI = confidence interval. RR = rate ratio.

• 

  Fig. 3. Odds ratios for unhealthy and moderately healthy lifestyle scores, in reference to healthy lifestyle score, for emotional distress and quality of life. Bars represent odds ratios with 95% confidence intervals, using logistic regressions. Results are from longitudinal data with generalized estimating equations. All analyses were adjusted for attained age as cubic splines, sex, race/ethnicity, age at cancer diagnosis, socioeconomic status (including education, marriage status, household income, and insurance), and treatment variables chosen from backward selection for each outcome. All statistical tests were two-sided. No correction was performed for multiple testing since all outcomes were distinct hypotheses of individual interest. Number of participants in the total population is 18,664, with analysis of each outcome differing slightly due to pre-existing conditions, as reported in Supplementary Table 1. Source data are provided as a Source Data file. CI = confidence interval. OR = odds ratio.

• 

  Fig. 4. Population attributable fractions for unfavorable lifestyle, radiotherapy, and chemotherapy for chronic health conditions. Bar chart illustrating the population attributable fractions (PAF) for unfavorable lifestyle, radiotherapy, and chemotherapy for A cardiometabolic disease, B cardiovascular disease, C other, and D emotional distress and impaired quality of life (QoL). Number of participants in the total population is 18,664, with analysis of each outcome differing slightly due to pre-existing conditions, as reported in Supplementary Table 1. Source data are provided as a Source Data file.

• 

  Fig. 5. Population attributable fractions for individual lifestyle factors and chronic health conditions. Bar chart illustrating the population attributable fractions (PAF) for heavy/risky drinking, smoking, unfavorable body mass index (BMI), and insufficient physical activity for (A) cardiometabolic disease, (B) cardiovascular disease, (C) other, and (D) emotional distress and impaired quality of life (QoL). Number of participants in the total population is 18,664, with analysis of each outcome differing slightly due to pre-existing conditions, as reported in Supplementary Table 1. Source data are provided as a Source Data file.

• 

  Fig. 6. Absolute excess risk per 10,000 person-years for developing chronic health conditions associated with unhealthy or moderately healthy lifestyle score compared to healthy lifestyle in childhood cancer survivors and sibling controls. Bars represent AER with 95% confidence intervals. Analyses are adjusted for attained age as cubic splines, sex, race/ethnicity, age at cancer diagnosis, socioeconomic status (including education, marriage status, household income, and insurance), and treatment variables chosen from backward selection for each outcome. Number of participants in the total population is 18,664, with analysis of each outcome differing slightly due to pre-existing conditions, as reported in Supplementary Table 1. Source data are provided as a Source Data file. AER = absolute excess risk. CI = confidence interval.

• 

  Fig. 7. Absolute excess risk per 10,000 person-years for developing chronic health conditions associated with unhealthy or moderately healthy lifestyle score compared to healthy lifestyle in childhood cancer survivors and sibling controls. Bars represent AER with 95% confidence intervals. Analyses are adjusted for attained age as cubic splines, sex, race/ethnicity, age at cancer diagnosis, socioeconomic status (including education, marriage status, household income, and insurance), and treatment variables chosen from backward selection for each outcome. Number of participants in the total population is 18,664, with analysis of each outcome differing slightly due to pre-existing conditions, as reported in Supplementary Table 1. Source data are provided as a Source Data file. AER = absolute excess risk. CI = confidence interval.

• 

  Fig. 8. Risk reduction (probability difference, per 100) for emotional distress and impaired quality of life associated with unhealthy or moderately healthy lifestyle score compared to healthy lifestyle in childhood cancer survivors and sibling controls. Bars represent Risk reduction with 95% confidence intervals. Analyses were adjusted for attained age as cubic splines, sex, and race/ethnicity. Number of participants in the total population is 18,664, with analysis of each outcome differing slightly due to pre-existing conditions, as reported in Supplementary Table 1. Source data are provided as a Source Data file. CI = confidence interval.