Djm. Malvy et al., SERUM BETA-CAROTENE AND ANTIOXIDANT MICRONUTRIENTS IN CHILDREN WITH CANCER, International journal of epidemiology, 22(5), 1993, pp. 761-771
Serum antioxidant vitamins A (retinol) and E (alpha-tocopherol), beta-
carotene, zinc and selenium for 418 children with newly diagnosed mali
gnancy were compared with those of 632 cancer-free controls. Incident
cancer cases and controls were 1-16 years old and recruited in 1986-19
89. Age- and sex-adjusted serum concentrations of retinol, beta-carote
ne and alpha-tocopherol were significantly inversely associated with c
ancer. In similar models, the odds ratio (OR) comparing the highest wi
th the lowest quintile was 2.06 (95% confidence interval [CI] 1.40-3.0
2) for retinol, 3.87 (95% CI : 2.54-5.90) for beta-carotene, 2.15 (95%
CI : 1.48-3.10) for alpha-tocopherol, 1.29 (95% CI : 0.75-2.23) for s
elenium, and 1.94 (95% CI : 1.17-2.23) for zinc. The cancer sites that
were associated with serum beta-carotene were, in general, leukaemia,
lymphoma, central nervous system, bone and renal tumours. Moreover, l
eukaemia was associated with low mean serum levels of retinol, seleniu
m and zinc. Subjects with lymphoma, bone and renal tumours also had lo
wer mean retinol and alpha-tocopherol levels than controls. Brain tumo
ur patients had low vitamin E levels. Low serum values of antioxidant
vitamins were associated with childhood neoplasm occurrence. Some site
-specific effect was reported. Low peripheral nutrient levels are not
considered as cancer promoters but rather as an impairment of the body
's defence mechanism occurring during the cancer-related metabolic and
nutritional disturbances and inflammation processes.