Results

Of 882 eligible survivors aged 16-19 years, 794 received the questionnaire (Figure 1 ). Of those, 543 completed the questionnaire (response rate 68%). We excluded 39 (5%) survivors with missing information on body image and finally included 504 (64%) in the analysis. The final sibling population consisted of 136 participants. Most survivors and siblings were from a German speaking region and did not report a migration background (Table 1 ). Thirty percent of survivors and siblings reported that they currently have a boyfriend or girlfriend. Among survivors, median age at diagnosis was 5.8 years (interquartile range [IQR] 2.6-9.9), median time since diagnosis 11.6 years (IQR 8.2-14.8), and median age at study 17.7 years (IQR 16.8-18.6). The most frequent cancer diagnoses were leukaemia (32%) and CNS tumours (17%). Clinical characteristics were similar between participating and non-participating survivors (Supplementary table S1 ).
Chronic health conditions were reported by 354 (70%) survivors with 141 (28%) reporting 1 condition and 213 (42%) ≥ 2 conditions. Neurological (33%), renal and digestive (23%), and musculoskeletal conditions (23%) were most frequently reported (Table 2 ). The most commonly reported symptoms within these conditions were balance disorders (15%), chronic constipation or diarrhoea (9%), and prolonged pain in bones or joints (9%). Growth hormone deficiency was reported by 49 (10%) of survivors.

Body image in adolescent childhood cancer survivors and their siblings

Survivors and siblings reported overall comparable levels of body image (Figure 2; all p from chi-squared tests and ordered logistic regression >0.05). Few survivors (4.6%) and siblings (2.9%) completely disagreed with the statement I am satisfied with my body image . Slightly more survivors than siblings completely agreed with the statements I would like to change a few things regarding my body (28.9% vs. 24.0%) and I would like to change many things regarding my body (7.6% vs. 2.5%).

Determinants of a more negative body image in survivors

In terms of socio-demographic determinants, we found that female survivors were more likely to have a negative body image compared to male survivors (Table 3 ). This was evident for all three body image statements (all ORs ≥ 1.7). After adjustment for age at study and sex, we further found that survivors from French or Italian-speaking Switzerland (OR=1.6, 95%-CI: 1.1-2.3) and those with a migration background (OR=1.5, 95%-CI: 1.0-2.1) more often would like to change many things regarding their bodies. We identified cancer treatment as the most important clinical determinant. Compared to surgery only, survivors that received HSCT were more likely to have a negative body image. This was observed for all three body image statements (all ORs ≥ 2.2). CNS tumour survivors were less often satisfied with their body image compared to leukaemia survivors (OR=1.6, 95%-CI: 1.0-2.7).
Survivors with ≥ 2 chronic health conditions were more likely to have a negative body image compared to survivors without chronic health conditions (Table 4 ). This was observed for all three body image statements (all ORs ≥ 1.5). Survivors suffering from musculoskeletal and endocrine conditions consistently reported a more negative body image compared to survivors without such conditions. A separate analysis of growth hormone deficiency revealed particularly pronounced associations. Survivors with growth hormone deficiency were less often satisfied with their body image (OR=3.4, 95%-CI: 1.9-6.0) and more often would like to change few (OR=2.2, 95%-CI: 1.3-3.9) or many things regarding their bodies (OR=2.3, 95%-CI: 1.3-4.1). Survivors with renal and digestive conditions more often would like to change few things regarding their bodies (OR=1.6, 95%-CI: 1.1-2.4). Survivors with neurological conditions were less often satisfied with their body image (OR=1.4, 95%-CI: 1.0-2.0). These associations remained virtually unchanged after adjustment for age at study and sex.