There is a lack of studies that analyze the interaction between risk variables as predictors of back pain (BP) in adolescents. The objectives of this study were to examine the relationship between BP and several risk variables, and to analyze the effect of the mediation of gender in this association. This cross-sectional study included n = 617 adolescents (mean age: 14.10 ± 1.18 years old) who completed the BP Adolescent Survey and who performed the bench trunk curl (BTC) and Sorensen (SOR) tests. Males showed a significantly lower prevalence (OR: 0.67) and frequency (contingency coefficient: 1.73) of BP than females, less participation in leisure-time sedentary behaviors (LRSBs) per day of more than 2 h (Cramer's V: 0.110), a higher sufficient Physical Activity (PA) (Cramer's V: 0.323) and a higher score in the BTC test (Contingency coefficient: 0.346). A high BTC score indicated significantly lower risk of BP than mid (OR: 1.74; p = 0.025) or low (OR: 1.62; p = 0.022) BTC. The mediation analysis showed a significant indirect effect with a significant value for the Sobel test (z = 7.45 ± 0.111). When the BTC test value was included in the equations, the connection between gender and BP was no longer significant. There was a difference in the prevalence between gender in BP and LRSB and PA. BP was associated with the SOR test. The association between BP and gender was mediated by SOR results.
Keywords: adolescent; back pain; exercise; low back pain; paraspinal muscles; physical fitness.