4.2 Pharmacological explanations for the association between SSRIs and risk of UGIB
There are potential pharmacological explanations for the dose-response relationship between fluoxetine use and UGIB risk. Fluoxetine was found to stimulate gastric acid secretion in a dose-response fashion, and larger amounts of gastric acid could cause formation of chronic ulcer[28, 29]. Furthermore, fluoxetine decreased the incidence of indomethacin (one kind of NSAID)-induced ulcers and ulcer bleeding in rats[30, 31]. However, in our analysis, the dose-response relationship between fluoxetine and UGIB remained significant after controlling for NSAID use and other potential confounders. Despite the fact that fluoxetine is not generally recommended for older patients due to its long half-life and prolonged side effects[32], it was the most frequently prescribed SSRI for both case and control groups in our analysis and in other studies[33].
Our observation showed that paroxetine was associated with decreased risk of UGIB, which was an unexpected result. Despite being known as the most potent inhibitor of serotonin reuptake[34], previous investigations did not find an association between paroxetine use and UGIB risk[15, 25, 35]. However, a previous study has revealed that the paroxetine effect on platelet function was modified by the serotonin transporter gene (SLC6A4) promoter polymorphism (5-HTTLPR)[36]. After receiving paroxetine treatment, an increase in bleeding time was observed in patients without L-alleles but not in those with L-alleles. Patients without L-alleles experienced problems in platelet function such as greater decreases in platelet serotonin levels[36]. In contrast, a study conducted in Taiwan demonstrated no significant differences between three 5-HTTLPR genotype groups[37]. More studies are needed to clarify the role of genetic profile.
Our findings, which emphasized the association between UGIB risk and recent use of fluoxetine, are consistent with those of a large systemic review and meta-analysis, which found a 2.35-fold increased odds of UGIB in patients using SSRIs for less than a month[38]. Meanwhile, no association was observed between the time window of exposure and other classes of SSRIs.