Table 2.

Association between statin use and gastric cancer risk (whole cohort and stratified analysis according to noncardia and cardia regions).

Statin useUnivariate analysis (n = 63,605, GC = 169)PS matchinga (n = 22,870, GC = 62)PS adjustmenta (n = 57,243, GC = 150)Multivariable analysis (n = 63,605, GC = 169)
All GCSHR (95% CI)P valueSHR (95% CI)P valueSHR (95% CI)P valueSHR (95% CI)P value
Nonstatin useRefRefRefRef
Statin use0.61 (0.41–0.92)0.0200.34 (0.19–0.61)<0.0010.33 (0.18–0.59)<0.0010.44 (0.28–0.68)<0.001
Noncardia GC(n = 63,571, GC = 135)(n = 22,865, GC = 36)(n = 57,123, GC = 120)(n = 63,571, GC = 135)
SHR (95% CI)P valueSHR (95% CI)P valueSHR (95% CI)P valueSHR (95% CI)P value
Nonstatin useRefRefRefRef
Statin use0.56 (0.35– 0.90)0.0170.48 (0.24– 0.98)0.0440.33 (0.17– 0.65)0.0010.46 (0.27– 0.74)0.002
Cardia GC(n = 63,470, GC = 34)(n = 22,865) GC = 15)(n = 57,123, GC = 30)(n = 63,470, GC = 34)
SHR (95% CI)P valueSHR (95% CI)P valueSHR (95% CI)P valueSHR (95% CI)P value
Nonstatin useRefRefRefRef
Statin use0.83 (0.39–1.90)0.660n.a.b (n.a.b)n.a.b0.31 (0.09– 1.03)0.055n.a.b (n.a.b)n.a.b
  • Note: Statin use was defined as use for more than 180 days.

  • Abbreviations: GC, gastric cancer; SHR, subdistribution hazard ratio.

  • aPS analysis was performed after trimming of the extreme PS strata (5th and 95th percentiles).

  • bSHR could not be calculated as the estimation procedure for fitting the subdistribution hazard model failed to converge.