Children and adolescents with leukemia are potentially at a high risk of developing vitamin D deficiency, due to limited physical activity and sunlight exposure, poor nutrition, chemotherapy, and its complications. The prevalence of vitamin D inadequacy in European pediatric cancer patients has been reported to be high. It is not known how many patients already have vitamin D deficiency at the time of diagnosis and whether vitamin D status at the time of diagnosis influences clinical outcome.

We aimed to investigate vitamin D status in children with leukemia at the time of diagnosis and explore possible factors (age, type of leukemia, gender, year and season of sampling) contributing to a low level of 25-hydroxyvitamin D (25-OHD). Furthermore, we aimed to investigate if vitamin D status at the time of diagnosis influences overall survival. We carried out a cross-sectional study including all 295 children (169 boys, 57.3%) aged <18 years who were diagnosed with leukemia in our institution between 1991 and 2016 and had a stored serum sample available from the time of diagnosis. All samples had been stored at -80C. We analysed serum 25-OHD and PTH with reagents from the same batch in January 2018; 25-OHD levels <25 nmol/L were considered deficient, 25-50 nmol/L insufficient, 50-75 nmol/L sufficient, and ≥75nmol/L optimal. Clinical data (sex, age, diagnosis, date of the diagnosis, overall survival) were collected from the Swedish Childhood Cancer Registry.

Altogether 295 children were included: 232 of them had acute lymphoblastic leukaemia (ALL), 52 acute myeloid leukaemia (AML), and 11 other types of leukemia (8 chronic myeloid leukaemia and 3 juvenile myelomonocytic leukaemia). Mean 25-OHD concentration was 60.7 nmol/L (SD 23.3). One third of the children (33.2%) had a subnormal 25-OHD level (6.4% had deficiency and 26.8% insufficiency), 39.7% were sufficient and 27.1% had an optimal level. There was a significant negative correlation between serum 25-OHD and PTH (p<0.001).

Season affected serum 25-OHD: it was lowest in the spring (55.2 nmol/L, SD 21.7) and highest in the summer (68.4 nmol/L, SD 19.6).

Multiple linear regression with unadjusted and adjusted analyses to explore the impact of age, diagnosis, gender, season, and time of sampling (calendar year) on 25-OHD level indicated that significant predictors of lower 25-OHD level were older age (p<0.001), sampling in the spring (p<0.001), sampling in more recent calendar year (p=0.001) and sampling in the winter (p=0.001).

When exploring the impact of 25-OHD on survival, we used Cox proportional hazard regression. In the whole cohort only the diagnosis and the age at diagnosis were significant. However, when the younger patients (≤ 6 year of age) were analysed separately, 25-OHD level <50 nmol/L at the time of diagnosis was associated with inferior overall survival independently of other factors (HR 3.05, p=0.03) as compared with those with 25-OHD ≥50 nmol/L. This patient group included 163 patients with 16 events.

Conclusion: Subnormal 25-OHD levels are common in pediatric patients with leukemia already at the time of diagnosis. In younger children with leukemia 25-OHD level <50 nmol/L is associated with inferior survival.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution