Mediation analysis of erythrocyte lipophilic index on the association between BMI and risk of oral cancer

Abstract

Abstract Aims To explore the relationship between the fatty acid lipophilic index (LI) of the erythrocyte membrane and oral cancer risk, as well as to evaluate the possibility of LI acting as a mediator of the association between body mass index (BMI) and oral cancer. Method Twenty-three fatty acids (FAs) of the erythrocyte membrane were measured using gas chromatography in 380 patients with oral cancer and 387 control subjects. The LI was calculated based on the FA proportion and FA melting points. The association of BMI and erythrocyte LI with oral cancer risk was analysed using logistic regression. The mediation effect of LI on the association between BMI and oral cancer risk was evaluated using mediation analysis. Results Among the control group, 46.0% were overweight or obese, which was significantly higher than that of oral cancer patients (29.5%). Significant differences in erythrocyte membrane saturated fatty acids (SFAs), monounsaturated fatty acids (MUFAs), and polyunsaturated fatty acids (PUFAs) were observed between the patient and control groups. The proportion of C18:1 n-9 from the MUFA family increased in oral cancer patients (12.67%) compared with controls (12.21%). While the total proportion of n-3 PUFAs decreased in oral cancer patients compared with controls, with C20:5 n-3 decreasing from 0.66 to 0.47%, and C22:6 n-3 decreasing from 5.82 to 4.86%. The LI was lower in the control participants (M = 27.6, IQR: 27.3–27.9) than in the oral cancer patients (M = 28.2, IQR: 27.9–28.5). BMI was inversely associated with oral cancer risk with a fully adjusted OR of 0.59 (95% CI: 0.43–0.83), while LI was positively associated with oral cancer risk with a fully adjusted OR of 1.99 (95% CI:1.36–2.94). LI explained 7% of the variance in the relationship between BMI and oral cancer risk. Conclusions The distribution of the FA profile in erythrocyte membranes differed between the oral cancer patients and the control group. The LI derived from the profile of FAs was positively associated with the risk of oral cancer, and the associations between BMI and oral cancer risk can be explained, at least in part, by LI.