Scandium and oral cancer: A case-control study based on propensity score matching

Background Limited data on potential association between scandium (Sc) and the risk of developing oral cancer are available from both domestic and international sources, and studies on the relationship between trace elements and tumors are mainly retrospective without covariance matching.

Objective This study explores the association of Sc with oral cancer risk based on propensity score matching (PSM).

Methods The subjects in this case-control study consisted of 390 oral cancer patients who were histologically confirmed in the First Affiliated Hospital of Fujian Medical University between January 2010 and November 2018 and 1 219 cancer-free subjects who were community residents and ordered health examination. A 1:1 PSM was used to match the case group to the healthy control group for selected potential factors such as demographic characteristics and lifestyles, and 300 patients with oral cancer and 300 healthy controls were selected for final analysis. Serum samples were digested by microwave and detected for Sc concentrations by inductively coupled plasma mass spectrometry. The distributions of demographic characteristics and lifestyles between the case and control groups were compared by chi-square test. The difference in serum Sc between the cases and controls was assessed by Wilcoxon rank sum test. The odds ratios (OR) and corresponding 95% confidence intervals (95% CI) for the association of serum Sc level with oral cancer risk using tertiles of serum Sc level of the control group were estimated by conditional logistic regression model. The association between Sc level and the risk of oral cancer was also analyzed by Cochran-Armitage trend test.

Results The median (P₂₅-P₇₅) of serum Sc concentration in the oral cancer patients5.56 (3.71-6.94) μg·L^-1 was lower than that in the controls7.21 (6.14-8.83) μg·L^-1 (P < 0.001). The results of conditional logistic regression model showed medium and high levels of serum Sc versus low level of serum Sc were associated with a decreased risk of oral cancer, and the adjusted ORs were 0.29 (95% CI:0.18-0.45) and 0.14 (95% CI:0.08-0.25), respectively. The risk of oral cancer was decreased with the increase of serum Sc level (P_trend < 0.001).

Conclusion A higher serum Sc level is associated with a lower risk of oral cancer.